id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2502.01956 | Shashank Sharma | Shashank Sharma, Janina Hoffmann, Vinay Namboodiri | DHP: Discrete Hierarchical Planning for Hierarchical Reinforcement Learning Agents | null | null | null | null | cs.RO cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Hierarchical Reinforcement Learning (HRL) agents often struggle with long-horizon visual planning due to their reliance on error-prone distance metrics. We propose Discrete Hierarchical Planning (DHP), a method that replaces continuous distance estimates with discrete reachability checks to evaluate subgoal feasibility. DHP recursively constructs tree-structured plans by decomposing long-term goals into sequences of simpler subtasks, using a novel advantage estimation strategy that inherently rewards shorter plans and generalizes beyond training depths. In addition, to address the data efficiency challenge, we introduce an exploration strategy that generates targeted training examples for the planning modules without needing expert data. Experiments in 25-room navigation environments demonstrate $100\%$ success rate (vs $82\%$ baseline) and $73$-step average episode length (vs $158$-step baseline). The method also generalizes to momentum-based control tasks and requires only $\log N$ steps for replanning. Theoretical analysis and ablations validate our design choices. | [
{
"version": "v1",
"created": "Tue, 04 Feb 2025 03:05:55 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:30:20 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Sharma",
"Shashank",
""
],
[
"Hoffmann",
"Janina",
""
],
[
"Namboodiri",
"Vinay",
""
]
] |
2502.02641 | Julia Bodensteiner | J. Bodensteiner, T. Shenar, H. Sana, N. Britavskiy, P. A. Crowther, N. Langer, D. J. Lennon, L. Mahy, L. R. Patrick, J. I. Villase\~nor, M. Abdul-Masih, D. M. Bowman, A. de Koter, S. E. de Mink, K. Deshmukh, M. Fabry, A. Gilkis, Y. G\"otberg, G. Holgado, R. G. Izzard, S. Janssens, V. M. Kalari, Z. Keszthelyi, J. Kub\'at, I. Mandel, G. Maravelias, L. M. Oskinova, D. Pauli, V. Ramachandran, D. F. Rocha, M. Renzo, A. A. C. Sander, F. R. N. Schneider, A. Schootemeijer, K. Sen, M. Stoop, S. Toonen, J. Th. van Loon, R. Valli, A. Vigna-G\'omez, J. S. Vink, C. Wang, and X.-T. Xu | Binarity at Low Metallicity (BLOeM) -- Multiplicity properties of Oe and Be stars | 12 pages (+appendix), 7 figures, accepted for publication in A&A | A&A 698, A38 (2025) | 10.1051/0004-6361/202452623 | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Rapidly rotating classical OBe stars have been proposed as the products of binary interactions, and the fraction of Be stars with compact companions implies that at least some are. However, to constrain the interaction physics spinning up the OBe stars, a large sample of homogeneously analysed OBe stars with well-determined binary characteristics and orbital parameters are required. We investigate the multiplicity properties of a sample of 18 Oe, 62 Be, and two Of?p stars observed within the BLOeM survey in the Small Magellanic Cloud. We analyse the first nine epochs of spectroscopic observations obtained over approximately three months in 2023. Radial velocities (RVs) of all stars are measured. Applying commonly-used binarity criteria we classify objects as binaries, binary candidates, and apparently single (RV stable) objects. We further inspect the spectra for double-lined spectroscopic binaries and cross-match with catalogues of X-ray sources and photometric binaries. We classify 14 OBe stars as binaries, and an additional 11 as binary candidates. The two Of?p stars are apparently single. Two more objects are most likely currently interacting binaries. Without those, the observed binary fraction for the OBe sample (78 stars) is f_OBe_obs=0.18+/-0.04 (f_obs_cand=0.32+/-0.05 including candidates). This fraction is less than half of that measured for OB stars in BLOeM. Combined with the lower fraction of SB2s, this suggests that OBe stars have indeed fundamentally different binary properties than OB stars. We find no evidence for OBe binaries with massive compact companions, in contrast to expectations from binary population synthesis. Our results support the binary scenario as an important formation channel for OBe stars, as post-interaction binaries may have been disrupted or the stripped companions of OBe stars are harder to detect. | [
{
"version": "v1",
"created": "Tue, 04 Feb 2025 19:00:01 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bodensteiner",
"J.",
""
],
[
"Shenar",
"T.",
""
],
[
"Sana",
"H.",
""
],
[
"Britavskiy",
"N.",
""
],
[
"Crowther",
"P. A.",
""
],
[
"Langer",
"N.",
""
],
[
"Lennon",
"D. J.",
""
],
[
"Mahy",
"L.",
""
],
[
"Patrick",
"L. R.",
""
],
[
"Villaseñor",
"J. I.",
""
],
[
"Abdul-Masih",
"M.",
""
],
[
"Bowman",
"D. M.",
""
],
[
"de Koter",
"A.",
""
],
[
"de Mink",
"S. E.",
""
],
[
"Deshmukh",
"K.",
""
],
[
"Fabry",
"M.",
""
],
[
"Gilkis",
"A.",
""
],
[
"Götberg",
"Y.",
""
],
[
"Holgado",
"G.",
""
],
[
"Izzard",
"R. G.",
""
],
[
"Janssens",
"S.",
""
],
[
"Kalari",
"V. M.",
""
],
[
"Keszthelyi",
"Z.",
""
],
[
"Kubát",
"J.",
""
],
[
"Mandel",
"I.",
""
],
[
"Maravelias",
"G.",
""
],
[
"Oskinova",
"L. M.",
""
],
[
"Pauli",
"D.",
""
],
[
"Ramachandran",
"V.",
""
],
[
"Rocha",
"D. F.",
""
],
[
"Renzo",
"M.",
""
],
[
"Sander",
"A. A. C.",
""
],
[
"Schneider",
"F. R. N.",
""
],
[
"Schootemeijer",
"A.",
""
],
[
"Sen",
"K.",
""
],
[
"Stoop",
"M.",
""
],
[
"Toonen",
"S.",
""
],
[
"van Loon",
"J. Th.",
""
],
[
"Valli",
"R.",
""
],
[
"Vigna-Gómez",
"A.",
""
],
[
"Vink",
"J. S.",
""
],
[
"Wang",
"C.",
""
],
[
"Xu",
"X. -T.",
""
]
] |
2502.02644 | Lee Patrick | L. R. Patrick, D. J. Lennon, F. Najarro, T. Shenar, J. Bodensteiner, H. Sana, P. A. Crowther, N. Britavskiy, N. Langer, A. Schootemeijer, C. J. Evans, L. Mahy, Y. G\"otberg, S. E. de Mink, F. R. N. Schneider, A. J. G. O'Grady, J. I. Villase\~nor, M. Bernini-Peron, D. M. Bowman, A. de Koter, K. Deshmukh, A. Gilkis, G. Gonz\'alez-Tor\`a, V. M. Kalari, Z. \~Keszthelyi, I. Mandel, A. Menon, M. Moe, L. M. Oskinova, D. Pauli, M. Renzo, A. A. C. Sander, K. Sen, M. Stoop, J. Th. van Loon, S. Toonen, F. Tramper, J. S. Vink, C. Wang | Binarity at LOw Metallicity (BLOeM): The multiplicity properties and evolution of BAF-type supergiants | Accepted in A&A | A&A 698, A39 (2025) | 10.1051/0004-6361/202452949 | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Given the uncertain evolutionary status of blue supergiant stars, their multiplicity properties hold vital clues to better understand their origin and evolution. As part of The Binarity at LOw Metallicity (BLOeM) campaign in the Small Magellanic Cloud we present a multi-epoch spectroscopic survey of 128 supergiant stars of spectral type B5--F5, which roughly correspond to initial masses in the range 6 to 30 solar masses. The observed binary fraction for the B5-9 supergiants is 25+/-6 % (10+/-4 %) and 5+/-2 % (0 %) for the A-F stars, using a radial velocity (RV) variability threshold of 5 kms (10 kms) as a criterion for binarity. Accounting for observational biases we find an intrinsic multiplicity fraction of less than 18% for the B5-9 stars and 8$^{+9}_{-7}$% for the AF stars, for the orbital periods up to 10$^{3.5}$day and mass-ratios (q) in the range 0.1 < q < 1. The large stellar radii of these supergiant stars prevent short orbital periods but we demonstrate that this effect alone cannot explain our results. We assess the spectra and RV time series of the detected binary systems and find that only a small fraction display convincing solutions. We conclude that the multiplicity fractions are compromised by intrinsic stellar variability such that the true multiplicity fraction may be significantly smaller. Our main conclusions from comparing the multiplicity properties of the B5-9 and AF supergiants to that of their less evolved counterparts is that such stars cannot be explained by a direct evolution from the main sequence. Furthermore, by comparing their multiplicity properties to red supergiant stars we conclude that the AF supergiant stars are neither progenitors nor descendants of red supergiants. | [
{
"version": "v1",
"created": "Tue, 04 Feb 2025 19:00:03 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Patrick",
"L. R.",
""
],
[
"Lennon",
"D. J.",
""
],
[
"Najarro",
"F.",
""
],
[
"Shenar",
"T.",
""
],
[
"Bodensteiner",
"J.",
""
],
[
"Sana",
"H.",
""
],
[
"Crowther",
"P. A.",
""
],
[
"Britavskiy",
"N.",
""
],
[
"Langer",
"N.",
""
],
[
"Schootemeijer",
"A.",
""
],
[
"Evans",
"C. J.",
""
],
[
"Mahy",
"L.",
""
],
[
"Götberg",
"Y.",
""
],
[
"de Mink",
"S. E.",
""
],
[
"Schneider",
"F. R. N.",
""
],
[
"O'Grady",
"A. J. G.",
""
],
[
"Villaseñor",
"J. I.",
""
],
[
"Bernini-Peron",
"M.",
""
],
[
"Bowman",
"D. M.",
""
],
[
"de Koter",
"A.",
""
],
[
"Deshmukh",
"K.",
""
],
[
"Gilkis",
"A.",
""
],
[
"González-Torà",
"G.",
""
],
[
"Kalari",
"V. M.",
""
],
[
"Keszthelyi",
"Z.",
""
],
[
"Mandel",
"I.",
""
],
[
"Menon",
"A.",
""
],
[
"Moe",
"M.",
""
],
[
"Oskinova",
"L. M.",
""
],
[
"Pauli",
"D.",
""
],
[
"Renzo",
"M.",
""
],
[
"Sander",
"A. A. C.",
""
],
[
"Sen",
"K.",
""
],
[
"Stoop",
"M.",
""
],
[
"van Loon",
"J. Th.",
""
],
[
"Toonen",
"S.",
""
],
[
"Tramper",
"F.",
""
],
[
"Vink",
"J. S.",
""
],
[
"Wang",
"C.",
""
]
] |
2502.03119 | Ricarda Graf | Ricarda Graf, Susan Todd, M. Fazil Baksh | Comparison of the Cox proportional hazards model and Random Survival Forest algorithm for predicting patient-specific survival probabilities in clinical trial data | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | The Cox proportional hazards model is often used to analyze data from Randomized Controlled Trials (RCT) with time-to-event outcomes. Random survival forest (RSF) is a machine-learning algorithm known for its high predictive performance. We conduct a comprehensive neutral comparison study to compare the performance of Cox regression and RSF in various simulation scenarios based on two reference datasets from RCTs. The motivation is to identify settings in which one method is preferable over the other when comparing different aspects of performance using measures according to the TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis) recommendations. Our results show that conclusions solely based on the C index, a performance measure that has been predominantly used in previous studies comparing predictive accuracy of the Cox-PH and RSF model based on real-world observational time-to-event data and that has been criticized by methodologists, may not be generalizable to other aspects of predictive performance. We found that measures of overall performance may generally give more reasonable results, and that the standard log-rank splitting rule used for the RSF may be outperformed by alternative splitting rules, in particular in nonproportional hazards settings. In our simulations, performance of the RSF suffers less in data with treatment-covariate interactions compared to data where these are absent. Performance of the Cox-PH model is affected by the violation of the proportional hazards assumption. | [
{
"version": "v1",
"created": "Wed, 05 Feb 2025 12:26:43 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:40:35 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Graf",
"Ricarda",
""
],
[
"Todd",
"Susan",
""
],
[
"Baksh",
"M. Fazil",
""
]
] |
2502.03193 | Sayyed Ahmad Abtahi | Sayyed Ahmad Abtahi, Mohsen Maddahali, Ahmad Bakhtafrouz | Spin photonic topological metasurface based on kagome lattice and leaky-wave application | IEEE Access (2025) | in IEEE Access, vol. 13, pp. 89811-89818, 2025 | 10.1109/ACCESS.2025.3567182 | null | physics.optics cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | The emerging field of topological metasurfaces offers unique advantages, particularly in robustness against backscattering in low-profile structures. The lattice configuration of these structures significantly influences the ability to achieve sharp turns in the propagation path. One of the most studied lattices in condensed matter physics is the kagome lattice, characterized by its hexagonal Brillouin zone, which displays a Dirac cone in its dispersion diagram. Previous research on kagome lattices in photonic topological insulators has primarily focused on valley types of insulators. This article introduces a spin topological metasurface based on the kagome lattice and its unit cell, enabling a broad range of sharp turns and propagation paths. The unit cell is compared to its hexagonal and 60-degree rhombic counterparts, and a parametric study of its dimensions is conducted. As a result of this research, a new X-band leaky-wave antenna designed in the kagome lattice with an armchair arrangement interface has been developed. This antenna provides two forward and two backward beams, each pair achieving an approximately 50-degree scan within the 8.8 to 11.1 GHz bandwidth. | [
{
"version": "v1",
"created": "Wed, 05 Feb 2025 14:12:53 GMT"
},
{
"version": "v2",
"created": "Tue, 13 May 2025 17:10:09 GMT"
},
{
"version": "v3",
"created": "Wed, 21 May 2025 18:30:37 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Abtahi",
"Sayyed Ahmad",
""
],
[
"Maddahali",
"Mohsen",
""
],
[
"Bakhtafrouz",
"Ahmad",
""
]
] |
2502.03536 | Sisi Zhou | Sisi Zhou, Senrui Chen | Randomized measurements for multi-parameter quantum metrology | 30 pages, 2 figures, numerical simulation added, presentation improved | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The optimal quantum measurements for estimating different unknown parameters in a parameterized quantum state are usually incompatible with each other. Traditional approaches to addressing the measurement incompatibility issue, such as the Holevo Cram\'{e}r--Rao bound, suffer from multiple difficulties towards practical applicability, as the optimal measurement strategies are usually state-dependent, difficult to implement and also take complex analyses to determine. Here we study randomized measurements as a new approach for multi-parameter quantum metrology. We show quantum measurements on single copies of quantum states given by $3$-design perform near-optimally when estimating an arbitrary number of parameters in pure states and more generally, approximately low-rank states, whose metrological information is largely concentrated in a low-dimensional subspace. The near-optimality is also shown in estimating the maximal number of parameters for three types of mixed states that are well-conditioned on its support. Examples of fidelity estimation and Hamiltonian estimation are explicitly provided to demonstrate the power and limitation of randomized measurements in multi-parameter quantum metrology. | [
{
"version": "v1",
"created": "Wed, 05 Feb 2025 19:00:04 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 03:46:10 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhou",
"Sisi",
""
],
[
"Chen",
"Senrui",
""
]
] |
2502.03771 | Luis Gaspar Schroeder | Luis Gaspar Schroeder, Aditya Desai, Alejandro Cuadron, Kyle Chu, Shu Liu, Mark Zhao, Stephan Krusche, Alfons Kemper, Matei Zaharia, Joseph E. Gonzalez | vCache: Verified Semantic Prompt Caching | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Semantic caches return cached LLM-generated responses for semantically similar prompts to reduce inference latency and cost. They embed cached prompts and store them alongside their response in a vector database. Embedding similarity metrics assign a numerical score to quantify the similarity between a request and its nearest neighbor prompt from the cache. Existing systems use the same static similarity threshold across all requests to determine whether two prompts can share similar responses. However, we observe that static thresholds do not give formal correctness guarantees, can result in unexpected error rates, and lead to suboptimal cache hit rates. This paper proposes vCache, the first verified semantic cache with user-defined error rate guarantees. It employs an online learning algorithm to estimate an optimal threshold for each cached prompt, enabling reliable cache responses without additional training. Our experiments show that vCache consistently meets the specified error bounds while outperforming state-of-the-art static-threshold and fine-tuned embedding baselines. We release the vCache implementation and benchmarks to support future research. | [
{
"version": "v1",
"created": "Thu, 06 Feb 2025 04:16:20 GMT"
},
{
"version": "v2",
"created": "Fri, 04 Apr 2025 16:51:15 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 04:15:22 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Schroeder",
"Luis Gaspar",
""
],
[
"Desai",
"Aditya",
""
],
[
"Cuadron",
"Alejandro",
""
],
[
"Chu",
"Kyle",
""
],
[
"Liu",
"Shu",
""
],
[
"Zhao",
"Mark",
""
],
[
"Krusche",
"Stephan",
""
],
[
"Kemper",
"Alfons",
""
],
[
"Zaharia",
"Matei",
""
],
[
"Gonzalez",
"Joseph E.",
""
]
] |
2502.03783 | Luohong Wu | Luohong Wu, Nicola A. Cavalcanti, Matthias Seibold, Giuseppe Loggia, Lisa Reissner, Jonas Hein, Silvan Beeler, Arnd Vieh\"ofer, Stephan Wirth, Lilian Calvet, Philipp F\"urnstahl | UltraBones100k: A reliable automated labeling method and large-scale dataset for ultrasound-based bone surface extraction | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Ultrasound-based bone surface segmentation is crucial in computer-assisted orthopedic surgery. However, ultrasound images have limitations, including a low signal-to-noise ratio, and acoustic shadowing, which make interpretation difficult. Existing deep learning models for bone segmentation rely primarily on costly manual labeling by experts, limiting dataset size and model generalizability. Additionally, the complexity of ultrasound physics and acoustic shadow makes the images difficult for humans to interpret, leading to incomplete labels in anechoic regions and limiting model performance. To advance ultrasound bone segmentation and establish effective model benchmarks, larger and higher-quality datasets are needed.
We propose a methodology for collecting ex-vivo ultrasound datasets with automatically generated bone labels, including anechoic regions. The proposed labels are derived by accurately superimposing tracked bone CT models onto the tracked ultrasound images. These initial labels are refined to account for ultrasound physics. A clinical evaluation is conducted by an expert physician specialized on orthopedic sonography to assess the quality of the generated bone labels. A neural network for bone segmentation is trained on the collected dataset and its predictions are compared to expert manual labels, evaluating accuracy, completeness, and F1-score.
We collected the largest known dataset of 100k ultrasound images of human lower limbs with bone labels, called UltraBones100k. A Wilcoxon signed-rank test with Bonferroni correction confirmed that the bone alignment after our method significantly improved the quality of bone labeling (p < 0.001). The model trained on UltraBones100k consistently outperforms manual labeling in all metrics, particularly in low-intensity regions (320% improvement in completeness at a distance threshold of 0.5 mm). | [
{
"version": "v1",
"created": "Thu, 06 Feb 2025 05:11:29 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Feb 2025 07:03:17 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 10:44:53 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Wu",
"Luohong",
""
],
[
"Cavalcanti",
"Nicola A.",
""
],
[
"Seibold",
"Matthias",
""
],
[
"Loggia",
"Giuseppe",
""
],
[
"Reissner",
"Lisa",
""
],
[
"Hein",
"Jonas",
""
],
[
"Beeler",
"Silvan",
""
],
[
"Viehöfer",
"Arnd",
""
],
[
"Wirth",
"Stephan",
""
],
[
"Calvet",
"Lilian",
""
],
[
"Fürnstahl",
"Philipp",
""
]
] |
2502.04583 | Jaemoo Choi | Jaemoo Choi, Jaewoong Choi, Dohyun Kwon | Overcoming Spurious Solutions in Semi-Dual Neural Optimal Transport: A Smoothing Approach for Learning the Optimal Transport Plan | 21 pages, 10 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | We address the convergence problem in learning the Optimal Transport (OT) map, where the OT Map refers to a map from one distribution to another while minimizing the transport cost. Semi-dual Neural OT, a widely used approach for learning OT Maps with neural networks, often generates spurious solutions that fail to transfer one distribution to another accurately. We identify a sufficient condition under which the max-min solution of Semi-dual Neural OT recovers the true OT Map. Moreover, to address cases when this sufficient condition is not satisfied, we propose a novel method, OTP, which learns both the OT Map and the Optimal Transport Plan, representing the optimal coupling between two distributions. Under sharp assumptions on the distributions, we prove that our model eliminates the spurious solution issue and correctly solves the OT problem. Our experiments show that the OTP model recovers the optimal transport map where existing methods fail and outperforms current OT-based models in image-to-image translation tasks. Notably, the OTP model can learn stochastic transport maps when deterministic OT Maps do not exist, such as one-to-many tasks like colorization. | [
{
"version": "v1",
"created": "Fri, 07 Feb 2025 00:37:12 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:11:28 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Choi",
"Jaemoo",
""
],
[
"Choi",
"Jaewoong",
""
],
[
"Kwon",
"Dohyun",
""
]
] |
2502.04592 | Yang Zhang | Yang Zhang, Wenbo Yang, Jun Wang, Qiang Ma, Jie Xiong | CAMEF: Causal-Augmented Multi-Modality Event-Driven Financial Forecasting by Integrating Time Series Patterns and Salient Macroeconomic Announcements | null | null | null | null | cs.LG cs.AI cs.CE cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurately forecasting the impact of macroeconomic events is critical for investors and policymakers. Salient events like monetary policy decisions and employment reports often trigger market movements by shaping expectations of economic growth and risk, thereby establishing causal relationships between events and market behavior. Existing forecasting methods typically focus either on textual analysis or time-series modeling, but fail to capture the multi-modal nature of financial markets and the causal relationship between events and price movements. To address these gaps, we propose CAMEF (Causal-Augmented Multi-Modality Event-Driven Financial Forecasting), a multi-modality framework that effectively integrates textual and time-series data with a causal learning mechanism and an LLM-based counterfactual event augmentation technique for causal-enhanced financial forecasting. Our contributions include: (1) a multi-modal framework that captures causal relationships between policy texts and historical price data; (2) a new financial dataset with six types of macroeconomic releases from 2008 to April 2024, and high-frequency real trading data for five key U.S. financial assets; and (3) an LLM-based counterfactual event augmentation strategy. We compare CAMEF to state-of-the-art transformer-based time-series and multi-modal baselines, and perform ablation studies to validate the effectiveness of the causal learning mechanism and event types. | [
{
"version": "v1",
"created": "Fri, 07 Feb 2025 00:55:25 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhang",
"Yang",
""
],
[
"Yang",
"Wenbo",
""
],
[
"Wang",
"Jun",
""
],
[
"Ma",
"Qiang",
""
],
[
"Xiong",
"Jie",
""
]
] |
2502.05159 | Parjanya Prashant | Parjanya Prajakta Prashant, Kaustubh Ponkshe, Babak Salimi | A Lightweight Method to Disrupt Memorized Sequences in LLM | 26 pages, 3 figures | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | As language models scale, their performance improves dramatically across a wide range of tasks, but so does their tendency to memorize and regurgitate parts of their training data verbatim. This tradeoff poses serious legal, ethical, and safety concerns, especially in real-world deployments. Existing mitigation techniques, such as differential privacy or model unlearning, often require retraining or access to internal weights making them impractical for most users. In this work, we introduce TokenSwap, a lightweight, post-hoc defense designed for realistic settings where the user can only access token-level outputs. Our key insight is that while large models are necessary for high task performance, small models (e.g., DistilGPT-2) are often sufficient to assign fluent, grammatically plausible probabilities to common function words - and crucially, they memorize far less. By selectively swapping token probabilities between models, TokenSwap preserves the capabilities of large models while reducing their propensity for verbatim reproduction. Evaluations on Pythia-6.9B and Llama-3-8B show up to a 10$\times$ drop in exact memorization with negligible task degradation. Our method offers a practical, accessible solution for mitigating memorized generation in deployed LLMs. | [
{
"version": "v1",
"created": "Fri, 07 Feb 2025 18:41:21 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:54:17 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Prashant",
"Parjanya Prajakta",
""
],
[
"Ponkshe",
"Kaustubh",
""
],
[
"Salimi",
"Babak",
""
]
] |
2502.05374 | Chongyu Fan | Chongyu Fan, Jinghan Jia, Yihua Zhang, Anil Ramakrishna, Mingyi Hong, Sijia Liu | Towards LLM Unlearning Resilient to Relearning Attacks: A Sharpness-Aware Minimization Perspective and Beyond | Accepted by ICML 2025 | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | The LLM unlearning technique has recently been introduced to comply with data regulations and address the safety and ethical concerns of LLMs by removing the undesired data-model influence. However, state-of-the-art unlearning methods face a critical vulnerability: they are susceptible to ``relearning'' the removed information from a small number of forget data points, known as relearning attacks. In this paper, we systematically investigate how to make unlearned models robust against such attacks. For the first time, we establish a connection between robust unlearning and sharpness-aware minimization (SAM) through a unified robust optimization framework, in an analogy to adversarial training designed to defend against adversarial attacks. Our analysis for SAM reveals that smoothness optimization plays a pivotal role in mitigating relearning attacks. Thus, we further explore diverse smoothing strategies to enhance unlearning robustness. Extensive experiments on benchmark datasets, including WMDP and MUSE, demonstrate that SAM and other smoothness optimization approaches consistently improve the resistance of LLM unlearning to relearning attacks. Notably, smoothness-enhanced unlearning also helps defend against (input-level) jailbreaking attacks, broadening our proposal's impact in robustifying LLM unlearning. Codes are available at https://github.com/OPTML-Group/Unlearn-Smooth. | [
{
"version": "v1",
"created": "Fri, 07 Feb 2025 23:03:55 GMT"
},
{
"version": "v2",
"created": "Sun, 23 Feb 2025 20:04:22 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Mar 2025 12:18:42 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 03:29:37 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Fan",
"Chongyu",
""
],
[
"Jia",
"Jinghan",
""
],
[
"Zhang",
"Yihua",
""
],
[
"Ramakrishna",
"Anil",
""
],
[
"Hong",
"Mingyi",
""
],
[
"Liu",
"Sijia",
""
]
] |
2502.06042 | Louis B\'ethune | Louis Bethune, David Grangier, Dan Busbridge, Eleonora Gualdoni, Marco Cuturi, Pierre Ablin | Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection | 19 pages, 15 figures, preprint | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | A widespread strategy to obtain a language model that performs well on a target domain is to finetune a pretrained model to perform unsupervised next-token prediction on data from that target domain. Finetuning presents two challenges: (i) if the amount of target data is limited, as in most practical applications, the model will quickly overfit, and (ii) the model will drift away from the original model, forgetting the pretraining data and the generic knowledge that comes with it. We aim to derive scaling laws that quantify these two phenomena for various target domains, amounts of available target data, and model scales. We measure the efficiency of injecting pretraining data into the finetuning data mixture to avoid forgetting and mitigate overfitting. A key practical takeaway from our study is that injecting as little as 1% of pretraining data in the finetuning data mixture prevents the model from forgetting the pretraining set. | [
{
"version": "v1",
"created": "Sun, 09 Feb 2025 21:44:27 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 18:48:37 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bethune",
"Louis",
""
],
[
"Grangier",
"David",
""
],
[
"Busbridge",
"Dan",
""
],
[
"Gualdoni",
"Eleonora",
""
],
[
"Cuturi",
"Marco",
""
],
[
"Ablin",
"Pierre",
""
]
] |
2502.06388 | Jin Pu | Yi Xiong, Jin Pu, Yi Ling, Guo-Ping Li, Gao-Ming Deng | Investigating the shadows of new regular black holes with a Minkowski core: Effects of spherical accretion and core type differences | 33 figures,2tables | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the shadows and optical appearances of a new type of regular black holes (BHs) with a Minkowski core under different spherical accretion. These BHs are constructed by modifying the Newtonian potential based on the minimum observable length in the Generalized Uncertainty Principle (GUP). They correspond one-to-one with the traditional regular BHs with a de-Sitter (dS) core (such as Bardeen/Hayward BHs), characterized by quantum gravity effect parameter ($\alpha_0$) and spacetime deformation factor ($n$). We find that the characteristic parameters give rise to some novel observable features. For these new BHs, the shadow radius, photon sphere radius, and total observed intensity increase with the increase of $\alpha_0$ but decrease with the increase of $n$. Under different spherical accretion, the shadow and photon sphere radius are identical, but the total observed intensity under the static spherical accretion is greater than that under the infalling spherical accretion. In addition, we find that these regular BHs with different cores show differences in shadows and optical appearances, especially under the static spherical accretion. Compared with Bardeen BH, the new BH has a smaller total observed intensity, dimmer maximum luminosity, and smaller shadow and photon sphere radius. The larger $\alpha_0$ leads to more significant differences, and a similar trend is also seen in the comparison with Hayward BH. Under the infalling spherical accretion, these regular BHs with different cores only have slight differences in total observed intensity, shadow and photon sphere radius, which become more obvious when $\alpha_0$ is relatively large. It suggests that the unique spacetime features of these regular BHs with different cores can be distinguished through astronomical observations. | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 12:16:26 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 05:55:24 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Xiong",
"Yi",
""
],
[
"Pu",
"Jin",
""
],
[
"Ling",
"Yi",
""
],
[
"Li",
"Guo-Ping",
""
],
[
"Deng",
"Gao-Ming",
""
]
] |
2502.06704 | Carlos O. Lousto | Ezequiel Zubieta, Ryan Missel, Susana B. Araujo Furlan, Carlos O. Lousto, Federico Garc\'ia, Santiago del Palacio, Guillermo Gancio, Jorge A. Combi, Linwei Wang | Study of the 2024 major Vela glitch at the Argentine Institute of Radioastronomy | 13 pages, 7 figures, 11 tables | A&A 698, A72 (2025) | 10.1051/0004-6361/202554098 | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report here on new results of the systematic monitoring of southern glitching pulsars at the Argentine Institute of Radioastronomy. In particular, we study in this work the new major glitch in the Vela pulsar (PSR J0835$-$4510) that occurred on 2024 April 29. We aim to thoroughly characterise the rotational behaviour of the Vela pulsar around its last major glitch and investigate the statistical properties of its individual pulses around the glitch. We characterise the rotational behaviour of the pulsar around the glitch through the pulsar timing technique. We measured the glitch parameters by fitting timing residuals to the data collected during the days surrounding the event. In addition, we study Vela individual pulses during the days of observation just before and after the glitch. We selected nine days of observations around the major glitch on 2024 April 29 and studied their statistical properties with the Self-Organizing Maps (SOM) technique. We used Variational AutoEncoder (VAE) reconstruction of the pulses to separate them clearly from the noise. We obtain a precise timing solution for the glitch. We find two recovery terms of $\sim 3~\mathrm{days}$ and $\sim 17~\mathrm{days}$. We find a correlation of high amplitude with narrower pulses while not finding notable qualitative systematic changes before and after the glitch. | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 17:32:12 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zubieta",
"Ezequiel",
""
],
[
"Missel",
"Ryan",
""
],
[
"Furlan",
"Susana B. Araujo",
""
],
[
"Lousto",
"Carlos O.",
""
],
[
"García",
"Federico",
""
],
[
"del Palacio",
"Santiago",
""
],
[
"Gancio",
"Guillermo",
""
],
[
"Combi",
"Jorge A.",
""
],
[
"Wang",
"Linwei",
""
]
] |
2502.06775 | Geyu Liang | Geyu Liang, Senne Michielssen and Salar Fattahi | Enhancing Performance of Explainable AI Models with Constrained Concept Refinement | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The trade-off between accuracy and interpretability has long been a challenge in machine learning (ML). This tension is particularly significant for emerging interpretable-by-design methods, which aim to redesign ML algorithms for trustworthy interpretability but often sacrifice accuracy in the process. In this paper, we address this gap by investigating the impact of deviations in concept representations-an essential component of interpretable models-on prediction performance and propose a novel framework to mitigate these effects. The framework builds on the principle of optimizing concept embeddings under constraints that preserve interpretability. Using a generative model as a test-bed, we rigorously prove that our algorithm achieves zero loss while progressively enhancing the interpretability of the resulting model. Additionally, we evaluate the practical performance of our proposed framework in generating explainable predictions for image classification tasks across various benchmarks. Compared to existing explainable methods, our approach not only improves prediction accuracy while preserving model interpretability across various large-scale benchmarks but also achieves this with significantly lower computational cost. | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 18:53:15 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 08:51:06 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Liang",
"Geyu",
""
],
[
"Michielssen",
"Senne",
""
],
[
"Fattahi",
"Salar",
""
]
] |
2502.06832 | Xu Zhang | Xu Zhang, Kaidi Xu, Ziqing Hu, Ren Wang | Optimizing Robustness and Accuracy in Mixture of Experts: A Dual-Model Approach | 15 pages, 7 figures, accepted by ICML 2025 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mixture of Experts (MoE) have shown remarkable success in leveraging specialized expert networks for complex machine learning tasks. However, their susceptibility to adversarial attacks presents a critical challenge for deployment in robust applications. This paper addresses the critical question of how to incorporate robustness into MoEs while maintaining high natural accuracy. We begin by analyzing the vulnerability of MoE components, finding that expert networks are notably more susceptible to adversarial attacks than the router. Based on this insight, we propose a targeted robust training technique that integrates a novel loss function to enhance the adversarial robustness of MoE, requiring only the robustification of one additional expert without compromising training or inference efficiency. Building on this, we introduce a dual-model strategy that linearly combines a standard MoE model with our robustified MoE model using a smoothing parameter. This approach allows for flexible control over the robustness-accuracy trade-off. We further provide theoretical foundations by deriving certified robustness bounds for both the single MoE and the dual-model. To push the boundaries of robustness and accuracy, we propose a novel joint training strategy JTDMoE for the dual-model. This joint training enhances both robustness and accuracy beyond what is achievable with separate models. Experimental results on CIFAR-10 and TinyImageNet datasets using ResNet18 and Vision Transformer (ViT) architectures demonstrate the effectiveness of our proposed methods. The code is publicly available at https://github.com/TIML-Group/Robust-MoE-Dual-Model. | [
{
"version": "v1",
"created": "Wed, 05 Feb 2025 20:45:52 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Feb 2025 05:30:33 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 15:29:53 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhang",
"Xu",
""
],
[
"Xu",
"Kaidi",
""
],
[
"Hu",
"Ziqing",
""
],
[
"Wang",
"Ren",
""
]
] |
2502.07266 | Yuyang Wu | Yuyang Wu, Yifei Wang, Ziyu Ye, Tianqi Du, Stefanie Jegelka, Yisen Wang | When More is Less: Understanding Chain-of-Thought Length in LLMs | null | null | null | null | cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large Language Models (LLMs) employ Chain-of-Thought (CoT) reasoning to deconstruct complex problems. While longer CoTs are often presumed superior, this paper challenges that notion, arguing that longer is not always better. Drawing on combined evidence from real-world observations, controlled experiments, and theoretical analysis, we demonstrate that task accuracy typically follows an inverted U-shaped curve with CoT length, where performance initially improves but eventually decreases as the number of CoT steps increases. With controlled experiments, we further uncover the scaling behaviors of the optimal CoT length: it increases with task difficulty but decreases with model capability, exposing an inherent simplicity bias where more capable models favor shorter, more efficient CoT reasoning. This bias is also evident in Reinforcement Learning (RL) training, where models gravitate towards shorter CoTs as their accuracy improves. To have a deep understanding of these dynamics, we establish a simple theoretical model that formally proves these phenomena, including the optimal length's scaling laws and the emergence of simplicity bias during RL. Guided by this framework, we demonstrate significant practical benefits from training with optimally-lengthed CoTs and employing length-aware filtering at inference. These findings offer both a principled understanding of the "overthinking" phenomenon and multiple practical guidelines for CoT calibration, enabling LLMs to achieve optimal reasoning performance with adaptive CoTs tailored to task complexity and model capability. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 05:28:59 GMT"
},
{
"version": "v2",
"created": "Sat, 24 May 2025 09:58:11 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:56:52 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Wu",
"Yuyang",
""
],
[
"Wang",
"Yifei",
""
],
[
"Ye",
"Ziyu",
""
],
[
"Du",
"Tianqi",
""
],
[
"Jegelka",
"Stefanie",
""
],
[
"Wang",
"Yisen",
""
]
] |
2502.07268 | Chih-Chun Chien | Xin Wang, Jia-Chen Tang, Xu-Yang Hou, Hao Guo, and Chih-Chun Chien | Mixed-state geometric phases of coherent and squeezed spin states | 13 pages, 6 figures, submitted | null | null | null | quant-ph cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Two mixed-state geometric phases, known as the Uhlmann phase and interferometric geometric phase (IGP), of spin coherent states (CSSs) and spin squeezed states (SSSs) are analyzed. Exact solutions and numerical results of selected examples are presented. For the $j = 3/2$ CSS, the Uhlmann phase exhibits finite-temperature topological phase transitions characterized by abrupt jumps. The IGP for the same state similarly shows discontinuous jumps as the temperature varies. In the case of the $j = 1$ one-axis SSS, both Uhlmann phase and IGP display discrete finite-temperature jumps. By contrast, the $j = 1$ two-axis SSS shows no such transitions because the Uhlmann phase and IGP both vary smoothly with temperature. We also briefly discuss potential realizations and simulations related to these phenomena in spin systems. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 05:32:27 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Apr 2025 01:21:06 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 15:12:18 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Wang",
"Xin",
""
],
[
"Tang",
"Jia-Chen",
""
],
[
"Hou",
"Xu-Yang",
""
],
[
"Guo",
"Hao",
""
],
[
"Chien",
"Chih-Chun",
""
]
] |
2502.07359 | Pengcheng Ai | Pengcheng Ai, Hongtao Qin, Xiangming Sun, Kaiwen Shang | Towards energy-insensitive and robust neutron/gamma classification: A learning-based frequency-domain parametric approach | 16 pages, 10 figures | null | 10.1016/j.net.2025.103667 | null | physics.ins-det physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neutron/gamma discrimination has been intensively researched in recent years, due to its unique scientific value and widespread applications. With the advancement of detection materials and algorithms, nowadays we can achieve fairly good discrimination. However, further improvements rely on better utilization of detector raw signals, especially energy-independent pulse characteristics. We begin by discussing why figure-of-merit (FoM) is not a comprehensive criterion for high-precision neutron/gamma discriminators, and proposing a new evaluation method based on adversarial sampling. Inspired by frequency-domain analysis in existing literature, parametric linear/nonlinear models with minimum complexity are created, upon the discrete spectrum, with tunable parameters just as neural networks. We train the models on an open-source neutron/gamma dataset (CLYC crystals with silicon photomultipliers) preprocessed by charge normalization to discover and exploit energy-independent features. The performance is evaluated on different sampling rates and noise levels, in comparison with the frequency classification index and conventional methods. The frequency-domain parametric models show higher accuracy and better adaptability to variations of data integrity than other discriminators. The proposed method is also promising for online inference on economical hardware and portable devices. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 08:29:48 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2025 02:32:53 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:50:19 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ai",
"Pengcheng",
""
],
[
"Qin",
"Hongtao",
""
],
[
"Sun",
"Xiangming",
""
],
[
"Shang",
"Kaiwen",
""
]
] |
2502.07406 | Tong Liu | BESIII Collaboration: M. Ablikim, M. N. Achasov, P. Adlarson, O. Afedulidis, X. C. Ai, R. Aliberti, A. Amoroso, Q. An, Y. Bai, O. Bakina, I. Balossino, Y. Ban, H.-R. Bao, V. Batozskaya, K. Begzsuren, N. Berger, M. Berlowski, M. Bertani, D. Bettoni, F. Bianchi, E. Bianco, A. Bortone, I. Boyko, R. A. Briere, A. Brueggemann, H. Cai, X. Cai, A. Calcaterra, G. F. Cao, N. Cao, S. A. Cetin, X. Y. Chai, J. F. Chang, G. R. Che, Y. Z. Che, G. Chelkov, C. Chen, C. H. Chen, Chao Chen, G. Chen, H. S. Chen, H. Y. Chen, M. L. Chen, S. J. Chen, S. L. Chen, S. M. Chen, T. Chen, X. R. Chen, X. T. Chen, Y. B. Chen, Y. Q. Chen, Z. J. Chen, S. K. Choi, G. Cibinetto, F. Cossio, J. J. Cui, H. L. Dai, J. P. Dai, A. Dbeyssi, R. E. de Boer, D. Dedovich, C. Q. Deng, Z. Y. Deng, A. Denig, I. Denysenko, M. Destefanis, F. De Mori, B. Ding, X. X. Ding, Y. Ding, Y. Ding, J. Dong, L. Y. Dong, M. Y. Dong, X. Dong, M. C. Du, S. X. Du, Y. Y. Duan, Z. H. Duan, P. Egorov, G. F. Fan, J. J. Fan, Y. H. Fan, J. Fang, J. Fang, S. S. Fang, W. X. Fang, Y. Q. Fang, R. Farinelli, L. Fava, F. Feldbauer, G. Felici, C. Q. Feng, J. H. Feng, Y. T. Feng, M. Fritsch, C. D. Fu, J. L. Fu, Y. W. Fu, H. Gao, X. B. Gao, Y. N. Gao, Y. N. Gao, Yang Gao, S. Garbolino, I. Garzia, P. T. Ge, Z. W. Ge, C. Geng, E. M. Gersabeck, A. Gilman, K. Goetzen, L. Gong, W. X. Gong, W. Gradl, S. Gramigna, M. Greco, M. H. Gu, Y. T. Gu, C. Y. Guan, A. Q. Guo, L. B. Guo, M. J. Guo, R. P. Guo, Y. P. Guo, A. Guskov, J. Gutierrez, K. L. Han, T. T. Han, F. Hanisch, X. Q. Hao, F. A. Harris, K. K. He, K. L. He, F. H. Heinsius, C. H. Heinz, Y. K. Heng, C. Herold, T. Holtmann, P. C. Hong, G. Y. Hou, X. T. Hou, Y. R. Hou, Z. L. Hou, B. Y. Hu, H. M. Hu, J. F. Hu, Q. P. Hu, S. L. Hu, T. Hu, Y. Hu, G. S. Huang, K. X. Huang, L. Q. Huang, P. Huang, X. T. Huang, Y. P. Huang, Y. S. Huang, T. Hussain, F. H\"olzken, N. H\"usken, N. in der Wiesche, J. Jackson, S. Janchiv, Q. Ji, Q. P. Ji, W. Ji, X. B. Ji, X. L. Ji, Y. Y. Ji, X. Q. Jia, Z. K. Jia, D. Jiang, H. B. Jiang, P. C. Jiang, S. S. Jiang, T. J. Jiang, X. S. Jiang, Y. Jiang, J. B. Jiao, J. K. Jiao, Z. Jiao, S. Jin, Y. Jin, M. Q. Jing, X. M. Jing, T. Johansson, S. Kabana, N. Kalantar-Nayestanaki, X. L. Kang, X. S. Kang, M. Kavatsyuk, B. C. Ke, V. Khachatryan, A. Khoukaz, R. Kiuchi, O. B. Kolcu, B. Kopf, M. Kuessner, X. Kui, N. Kumar, A. Kupsc, W. K\"uhn, W. N. Lan, T. T. Lei, Z. H. Lei, M. Lellmann, T. Lenz, C. Li, C. Li, C. H. Li, Cheng Li, D. M. Li, F. Li, G. Li, H. B. Li, H. J. Li, H. N. Li, Hui Li, J. R. Li, J. S. Li, K. Li, K. L. Li, L. J. Li, Lei Li, M. H. Li, P. L. Li, P. R. Li, Q. M. Li, Q. X. Li, R. Li, T. Li, T. Y. Li, W. D. Li, W. G. Li, X. Li, X. H. Li, X. L. Li, X. Y. Li, X. Z. Li, Y. Li, Y. G. Li, Z. J. Li, Z. Y. Li, C. Liang, H. Liang, Y. F. Liang, Y. T. Liang, G. R. Liao, Y. P. Liao, J. Libby, A. Limphirat, C. C. Lin, C. X. Lin, D. X. Lin, T. Lin, B. J. Liu, B. X. Liu, C. Liu, C. X. Liu, F. Liu, F. H. Liu, Feng Liu, G. M. Liu, H. Liu, H. B. Liu, H. H. Liu, H. M. Liu, Huihui Liu, J. B. Liu, K. Liu, K. Y. Liu, Ke Liu, L. Liu, L. C. Liu, Lu Liu, M. H. Liu, P. L. Liu, Q. Liu, S. B. Liu, T. Liu, W. K. Liu, W. M. Liu, X. Liu, X. Liu, Y. Liu, Y. Liu, Y. B. Liu, Z. A. Liu, Z. D. Liu, Z. Q. Liu, X. C. Lou, F. X. Lu, H. J. Lu, J. G. Lu, Y. Lu, Y. P. Lu, Z. H. Lu, C. L. Luo, J. R. Luo, M. X. Luo, T. Luo, X. L. Luo, X. R. Lyu, Y. F. Lyu, F. C. Ma, H. Ma, H. L. Ma, J. L. Ma, L. L. Ma, L. R. Ma, Q. M. Ma, R. Q. Ma, R. Y. Ma, T. Ma, X. T. Ma, X. Y. Ma, Y. M. Ma, F. E. Maas, I. MacKay, M. Maggiora, S. Malde, Y. J. Mao, Z. P. Mao, S. Marcello, Y. H. Meng, Z. X. Meng, J. G. Messchendorp, G. Mezzadri, H. Miao, T. J. Min, R. E. Mitchell, X. H. Mo, B. Moses, N. Yu. Muchnoi, J. Muskalla, Y. Nefedov, F. Nerling, L. S. Nie, I. B. Nikolaev, Z. Ning, S. Nisar, Q. L. Niu, W. D. Niu, Y. Niu, S. L. Olsen, Q. Ouyang, S. Pacetti, X. Pan, Y. Pan, A. Pathak, Y. P. Pei, M. Pelizaeus, H. P. Peng, Y. Y. Peng, K. Peters, J. L. Ping, R. G. Ping, S. Plura, V. Prasad, F. Z. Qi, H. R. Qi, M. Qi, S. Qian, W. B. Qian, C. F. Qiao, J. H. Qiao, J. J. Qin, L. Q. Qin, L. Y. Qin, X. P. Qin, X. S. Qin, Z. H. Qin, J. F. Qiu, Z. H. Qu, C. F. Redmer, K. J. Ren, A. Rivetti, M. Rolo, G. Rong, Ch. Rosner, M. Q. Ruan, S. N. Ruan, N. Salone, A. Sarantsev, Y. Schelhaas, K. Schoenning, M. Scodeggio, K. Y. Shan, W. Shan, X. Y. Shan, Z. J. Shang, J. F. Shangguan, L. G. Shao, M. Shao, C. P. Shen, H. F. Shen, W. H. Shen, X. Y. Shen, B. A. Shi, H. Shi, J. L. Shi, J. Y. Shi, S. Y. Shi, X. Shi, J. J. Song, T. Z. Song, W. M. Song, Y. J. Song, Y. X. Song, S. Sosio, S. Spataro, F. Stieler, S. S Su, Y. J. Su, G. B. Sun, G. X. Sun, H. Sun, H. K. Sun, J. F. Sun, K. Sun, L. Sun, S. S. Sun, T. Sun, Y. J. Sun, Y. Z. Sun, Z. Q. Sun, Z. T. Sun, C. J. Tang, G. Y. Tang, J. Tang, M. Tang, Y. A. Tang, L. Y. Tao, M. Tat, J. X. Teng, V. Thoren, W. H. Tian, Y. Tian, Z. F. Tian, I. Uman, Y. Wan, S. J. Wang, B. Wang, Bo Wang, C. Wang, D. Y. Wang, H. J. Wang, J. J. Wang, J. P. Wang, K. Wang, L. L. Wang, L. W. Wang, M. Wang, N. Y. Wang, S. Wang, S. Wang, T. Wang, T. J. Wang, W. Wang, W. Wang, W. P. Wang, X. Wang, X. F. Wang, X. J. Wang, X. L. Wang, X. N. Wang, Y. Wang, Y. D. Wang, Y. F. Wang, Y. H. Wang, Y. L. Wang, Y. N. Wang, Y. Q. Wang, Yaqian Wang, Yi Wang, Z. Wang, Z. L. Wang, Z. Y. Wang, D. H. Wei, F. Weidner, S. P. Wen, Y. R. Wen, U. Wiedner, G. Wilkinson, M. Wolke, L. Wollenberg, C. Wu, J. F. Wu, L. H. Wu, L. J. Wu, Lianjie Wu, X. Wu, X. H. Wu, Y. H. Wu, Y. J. Wu, Z. Wu, L. Xia, X. M. Xian, B. H. Xiang, T. Xiang, D. Xiao, G. Y. Xiao, H. Xiao, Y. L. Xiao, Z. J. Xiao, C. Xie, X. H. Xie, Y. Xie, Y. G. Xie, Y. H. Xie, Z. P. Xie, T. Y. Xing, C. F. Xu, C. J. Xu, G. F. Xu, M. Xu, Q. J. Xu, Q. N. Xu, W. L. Xu, X. P. Xu, Y. Xu, Y. C. Xu, Z. S. Xu, F. Yan, L. Yan, W. B. Yan, W. C. Yan, W. P. Yan, X. Q. Yan, H. J. Yang, H. L. Yang, H. X. Yang, J. H. Yang, R. J. Yang, T. Yang, Y. Yang, Y. F. Yang, Y. X. Yang, Y. Z. Yang, Z. W. Yang, Z. P. Yao, M. Ye, M. H. Ye, Junhao Yin, Z. Y. You, B. X. Yu, C. X. Yu, G. Yu, J. S. Yu, M. C. Yu, T. Yu, X. D. Yu, C. Z. Yuan, J. Yuan, J. Yuan, L. Yuan, S. C. Yuan, Y. Yuan, Z. Y. Yuan, C. X. Yue, Ying Yue, A. A. Zafar, F. R. Zeng, S. H. Zeng, X. Zeng, Y. Zeng, Y. J. Zeng, Y. J. Zeng, X. Y. Zhai, Y. C. Zhai, Y. H. Zhan, A. Q. Zhang, B. L. Zhang, B. X. Zhang, D. H. Zhang, G. Y. Zhang, H. Zhang, H. Zhang, H. C. Zhang, H. H. Zhang, H. Q. Zhang, H. R. Zhang, H. Y. Zhang, J. Zhang, J. Zhang, J. J. Zhang, J. L. Zhang, J. Q. Zhang, J. S. Zhang, J. W. Zhang, J. X. Zhang, J. Y. Zhang, J. Z. Zhang, Jianyu Zhang, L. M. Zhang, Lei Zhang, P. Zhang, Q. Zhang, Q. Y. Zhang, R. Y. Zhang, S. H. Zhang, Shulei Zhang, X. M. Zhang, X. Y Zhang, X. Y. Zhang, Y. Zhang, Y. Zhang, Y. T. Zhang, Y. H. Zhang, Y. M. Zhang, Yan Zhang, Z. D. Zhang, Z. H. Zhang, Z. L. Zhang, Z. X. Zhang, Z. Y. Zhang, Z. Y. Zhang, Z. Z. Zhang, Zh. Zh. Zhang, G. Zhao, J. Y. Zhao, J. Z. Zhao, L. Zhao, Lei Zhao, M. G. Zhao, N. Zhao, R. P. Zhao, S. J. Zhao, Y. B. Zhao, Y. X. Zhao, Z. G. Zhao, A. Zhemchugov, B. Zheng, B. M. Zheng, J. P. Zheng, W. J. Zheng, X. R. Zheng, Y. H. Zheng, B. Zhong, X. Zhong, H. Zhou, J. Y. Zhou, S. Zhou, X. Zhou, X. K. Zhou, X. R. Zhou, X. Y. Zhou, Y. Z. Zhou, Z. C. Zhou, A. N. Zhu, J. Zhu, K. Zhu, K. J. Zhu, K. S. Zhu, L. Zhu, L. X. Zhu, S. H. Zhu, T. J. Zhu, W. D. Zhu, W. J. Zhu, W. Z. Zhu, Y. C. Zhu, Z. A. Zhu, J. H. Zou, J. Zu | Search for $e^+e^-\to K_S^0 K_S^0 h_c$ | null | null | 10.1007/JHEP05(2025)144 | null | hep-ex | http://creativecommons.org/licenses/by/4.0/ | Using $e^+e^-$ collision data at 13 center-of-mass energies ranging from 4.600 to 4.950 GeV collected with the BESIII detector, we search for the unmeasured $e^+e^-\to K_S^0 K_S^0 h_c$ process . No significant signal is observed, and the upper limits of the Born cross sections at each center-of-mass energy are presented. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 09:39:52 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:10:01 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"BESIII Collaboration",
"",
""
],
[
"Ablikim",
"M.",
""
],
[
"Achasov",
"M. N.",
""
],
[
"Adlarson",
"P.",
""
],
[
"Afedulidis",
"O.",
""
],
[
"Ai",
"X. C.",
""
],
[
"Aliberti",
"R.",
""
],
[
"Amoroso",
"A.",
""
],
[
"An",
"Q.",
""
],
[
"Bai",
"Y.",
""
],
[
"Bakina",
"O.",
""
],
[
"Balossino",
"I.",
""
],
[
"Ban",
"Y.",
""
],
[
"Bao",
"H. -R.",
""
],
[
"Batozskaya",
"V.",
""
],
[
"Begzsuren",
"K.",
""
],
[
"Berger",
"N.",
""
],
[
"Berlowski",
"M.",
""
],
[
"Bertani",
"M.",
""
],
[
"Bettoni",
"D.",
""
],
[
"Bianchi",
"F.",
""
],
[
"Bianco",
"E.",
""
],
[
"Bortone",
"A.",
""
],
[
"Boyko",
"I.",
""
],
[
"Briere",
"R. A.",
""
],
[
"Brueggemann",
"A.",
""
],
[
"Cai",
"H.",
""
],
[
"Cai",
"X.",
""
],
[
"Calcaterra",
"A.",
""
],
[
"Cao",
"G. F.",
""
],
[
"Cao",
"N.",
""
],
[
"Cetin",
"S. A.",
""
],
[
"Chai",
"X. Y.",
""
],
[
"Chang",
"J. F.",
""
],
[
"Che",
"G. R.",
""
],
[
"Che",
"Y. Z.",
""
],
[
"Chelkov",
"G.",
""
],
[
"Chen",
"C.",
""
],
[
"Chen",
"C. H.",
""
],
[
"Chen",
"Chao",
""
],
[
"Chen",
"G.",
""
],
[
"Chen",
"H. S.",
""
],
[
"Chen",
"H. Y.",
""
],
[
"Chen",
"M. L.",
""
],
[
"Chen",
"S. J.",
""
],
[
"Chen",
"S. L.",
""
],
[
"Chen",
"S. M.",
""
],
[
"Chen",
"T.",
""
],
[
"Chen",
"X. R.",
""
],
[
"Chen",
"X. T.",
""
],
[
"Chen",
"Y. B.",
""
],
[
"Chen",
"Y. Q.",
""
],
[
"Chen",
"Z. J.",
""
],
[
"Choi",
"S. K.",
""
],
[
"Cibinetto",
"G.",
""
],
[
"Cossio",
"F.",
""
],
[
"Cui",
"J. J.",
""
],
[
"Dai",
"H. L.",
""
],
[
"Dai",
"J. P.",
""
],
[
"Dbeyssi",
"A.",
""
],
[
"de Boer",
"R. E.",
""
],
[
"Dedovich",
"D.",
""
],
[
"Deng",
"C. Q.",
""
],
[
"Deng",
"Z. Y.",
""
],
[
"Denig",
"A.",
""
],
[
"Denysenko",
"I.",
""
],
[
"Destefanis",
"M.",
""
],
[
"De Mori",
"F.",
""
],
[
"Ding",
"B.",
""
],
[
"Ding",
"X. X.",
""
],
[
"Ding",
"Y.",
""
],
[
"Ding",
"Y.",
""
],
[
"Dong",
"J.",
""
],
[
"Dong",
"L. Y.",
""
],
[
"Dong",
"M. Y.",
""
],
[
"Dong",
"X.",
""
],
[
"Du",
"M. C.",
""
],
[
"Du",
"S. X.",
""
],
[
"Duan",
"Y. Y.",
""
],
[
"Duan",
"Z. H.",
""
],
[
"Egorov",
"P.",
""
],
[
"Fan",
"G. F.",
""
],
[
"Fan",
"J. J.",
""
],
[
"Fan",
"Y. H.",
""
],
[
"Fang",
"J.",
""
],
[
"Fang",
"J.",
""
],
[
"Fang",
"S. S.",
""
],
[
"Fang",
"W. X.",
""
],
[
"Fang",
"Y. Q.",
""
],
[
"Farinelli",
"R.",
""
],
[
"Fava",
"L.",
""
],
[
"Feldbauer",
"F.",
""
],
[
"Felici",
"G.",
""
],
[
"Feng",
"C. Q.",
""
],
[
"Feng",
"J. H.",
""
],
[
"Feng",
"Y. T.",
""
],
[
"Fritsch",
"M.",
""
],
[
"Fu",
"C. D.",
""
],
[
"Fu",
"J. L.",
""
],
[
"Fu",
"Y. W.",
""
],
[
"Gao",
"H.",
""
],
[
"Gao",
"X. B.",
""
],
[
"Gao",
"Y. N.",
""
],
[
"Gao",
"Y. N.",
""
],
[
"Gao",
"Yang",
""
],
[
"Garbolino",
"S.",
""
],
[
"Garzia",
"I.",
""
],
[
"Ge",
"P. T.",
""
],
[
"Ge",
"Z. W.",
""
],
[
"Geng",
"C.",
""
],
[
"Gersabeck",
"E. M.",
""
],
[
"Gilman",
"A.",
""
],
[
"Goetzen",
"K.",
""
],
[
"Gong",
"L.",
""
],
[
"Gong",
"W. X.",
""
],
[
"Gradl",
"W.",
""
],
[
"Gramigna",
"S.",
""
],
[
"Greco",
"M.",
""
],
[
"Gu",
"M. H.",
""
],
[
"Gu",
"Y. T.",
""
],
[
"Guan",
"C. Y.",
""
],
[
"Guo",
"A. Q.",
""
],
[
"Guo",
"L. B.",
""
],
[
"Guo",
"M. J.",
""
],
[
"Guo",
"R. P.",
""
],
[
"Guo",
"Y. P.",
""
],
[
"Guskov",
"A.",
""
],
[
"Gutierrez",
"J.",
""
],
[
"Han",
"K. L.",
""
],
[
"Han",
"T. T.",
""
],
[
"Hanisch",
"F.",
""
],
[
"Hao",
"X. Q.",
""
],
[
"Harris",
"F. A.",
""
],
[
"He",
"K. K.",
""
],
[
"He",
"K. L.",
""
],
[
"Heinsius",
"F. H.",
""
],
[
"Heinz",
"C. H.",
""
],
[
"Heng",
"Y. K.",
""
],
[
"Herold",
"C.",
""
],
[
"Holtmann",
"T.",
""
],
[
"Hong",
"P. C.",
""
],
[
"Hou",
"G. Y.",
""
],
[
"Hou",
"X. T.",
""
],
[
"Hou",
"Y. R.",
""
],
[
"Hou",
"Z. L.",
""
],
[
"Hu",
"B. Y.",
""
],
[
"Hu",
"H. M.",
""
],
[
"Hu",
"J. F.",
""
],
[
"Hu",
"Q. P.",
""
],
[
"Hu",
"S. L.",
""
],
[
"Hu",
"T.",
""
],
[
"Hu",
"Y.",
""
],
[
"Huang",
"G. S.",
""
],
[
"Huang",
"K. X.",
""
],
[
"Huang",
"L. Q.",
""
],
[
"Huang",
"P.",
""
],
[
"Huang",
"X. T.",
""
],
[
"Huang",
"Y. P.",
""
],
[
"Huang",
"Y. S.",
""
],
[
"Hussain",
"T.",
""
],
[
"Hölzken",
"F.",
""
],
[
"Hüsken",
"N.",
""
],
[
"der Wiesche",
"N. in",
""
],
[
"Jackson",
"J.",
""
],
[
"Janchiv",
"S.",
""
],
[
"Ji",
"Q.",
""
],
[
"Ji",
"Q. P.",
""
],
[
"Ji",
"W.",
""
],
[
"Ji",
"X. B.",
""
],
[
"Ji",
"X. L.",
""
],
[
"Ji",
"Y. Y.",
""
],
[
"Jia",
"X. Q.",
""
],
[
"Jia",
"Z. K.",
""
],
[
"Jiang",
"D.",
""
],
[
"Jiang",
"H. B.",
""
],
[
"Jiang",
"P. C.",
""
],
[
"Jiang",
"S. S.",
""
],
[
"Jiang",
"T. J.",
""
],
[
"Jiang",
"X. S.",
""
],
[
"Jiang",
"Y.",
""
],
[
"Jiao",
"J. B.",
""
],
[
"Jiao",
"J. K.",
""
],
[
"Jiao",
"Z.",
""
],
[
"Jin",
"S.",
""
],
[
"Jin",
"Y.",
""
],
[
"Jing",
"M. Q.",
""
],
[
"Jing",
"X. M.",
""
],
[
"Johansson",
"T.",
""
],
[
"Kabana",
"S.",
""
],
[
"Kalantar-Nayestanaki",
"N.",
""
],
[
"Kang",
"X. L.",
""
],
[
"Kang",
"X. S.",
""
],
[
"Kavatsyuk",
"M.",
""
],
[
"Ke",
"B. C.",
""
],
[
"Khachatryan",
"V.",
""
],
[
"Khoukaz",
"A.",
""
],
[
"Kiuchi",
"R.",
""
],
[
"Kolcu",
"O. B.",
""
],
[
"Kopf",
"B.",
""
],
[
"Kuessner",
"M.",
""
],
[
"Kui",
"X.",
""
],
[
"Kumar",
"N.",
""
],
[
"Kupsc",
"A.",
""
],
[
"Kühn",
"W.",
""
],
[
"Lan",
"W. N.",
""
],
[
"Lei",
"T. T.",
""
],
[
"Lei",
"Z. H.",
""
],
[
"Lellmann",
"M.",
""
],
[
"Lenz",
"T.",
""
],
[
"Li",
"C.",
""
],
[
"Li",
"C.",
""
],
[
"Li",
"C. H.",
""
],
[
"Li",
"Cheng",
""
],
[
"Li",
"D. M.",
""
],
[
"Li",
"F.",
""
],
[
"Li",
"G.",
""
],
[
"Li",
"H. B.",
""
],
[
"Li",
"H. J.",
""
],
[
"Li",
"H. N.",
""
],
[
"Li",
"Hui",
""
],
[
"Li",
"J. R.",
""
],
[
"Li",
"J. S.",
""
],
[
"Li",
"K.",
""
],
[
"Li",
"K. L.",
""
],
[
"Li",
"L. J.",
""
],
[
"Li",
"Lei",
""
],
[
"Li",
"M. H.",
""
],
[
"Li",
"P. L.",
""
],
[
"Li",
"P. R.",
""
],
[
"Li",
"Q. M.",
""
],
[
"Li",
"Q. X.",
""
],
[
"Li",
"R.",
""
],
[
"Li",
"T.",
""
],
[
"Li",
"T. Y.",
""
],
[
"Li",
"W. D.",
""
],
[
"Li",
"W. G.",
""
],
[
"Li",
"X.",
""
],
[
"Li",
"X. H.",
""
],
[
"Li",
"X. L.",
""
],
[
"Li",
"X. Y.",
""
],
[
"Li",
"X. Z.",
""
],
[
"Li",
"Y.",
""
],
[
"Li",
"Y. G.",
""
],
[
"Li",
"Z. J.",
""
],
[
"Li",
"Z. Y.",
""
],
[
"Liang",
"C.",
""
],
[
"Liang",
"H.",
""
],
[
"Liang",
"Y. F.",
""
],
[
"Liang",
"Y. T.",
""
],
[
"Liao",
"G. R.",
""
],
[
"Liao",
"Y. P.",
""
],
[
"Libby",
"J.",
""
],
[
"Limphirat",
"A.",
""
],
[
"Lin",
"C. C.",
""
],
[
"Lin",
"C. X.",
""
],
[
"Lin",
"D. X.",
""
],
[
"Lin",
"T.",
""
],
[
"Liu",
"B. J.",
""
],
[
"Liu",
"B. X.",
""
],
[
"Liu",
"C.",
""
],
[
"Liu",
"C. X.",
""
],
[
"Liu",
"F.",
""
],
[
"Liu",
"F. H.",
""
],
[
"Liu",
"Feng",
""
],
[
"Liu",
"G. M.",
""
],
[
"Liu",
"H.",
""
],
[
"Liu",
"H. B.",
""
],
[
"Liu",
"H. H.",
""
],
[
"Liu",
"H. M.",
""
],
[
"Liu",
"Huihui",
""
],
[
"Liu",
"J. B.",
""
],
[
"Liu",
"K.",
""
],
[
"Liu",
"K. Y.",
""
],
[
"Liu",
"Ke",
""
],
[
"Liu",
"L.",
""
],
[
"Liu",
"L. C.",
""
],
[
"Liu",
"Lu",
""
],
[
"Liu",
"M. H.",
""
],
[
"Liu",
"P. L.",
""
],
[
"Liu",
"Q.",
""
],
[
"Liu",
"S. B.",
""
],
[
"Liu",
"T.",
""
],
[
"Liu",
"W. K.",
""
],
[
"Liu",
"W. M.",
""
],
[
"Liu",
"X.",
""
],
[
"Liu",
"X.",
""
],
[
"Liu",
"Y.",
""
],
[
"Liu",
"Y.",
""
],
[
"Liu",
"Y. B.",
""
],
[
"Liu",
"Z. A.",
""
],
[
"Liu",
"Z. D.",
""
],
[
"Liu",
"Z. Q.",
""
],
[
"Lou",
"X. C.",
""
],
[
"Lu",
"F. X.",
""
],
[
"Lu",
"H. J.",
""
],
[
"Lu",
"J. G.",
""
],
[
"Lu",
"Y.",
""
],
[
"Lu",
"Y. P.",
""
],
[
"Lu",
"Z. H.",
""
],
[
"Luo",
"C. L.",
""
],
[
"Luo",
"J. R.",
""
],
[
"Luo",
"M. X.",
""
],
[
"Luo",
"T.",
""
],
[
"Luo",
"X. L.",
""
],
[
"Lyu",
"X. R.",
""
],
[
"Lyu",
"Y. F.",
""
],
[
"Ma",
"F. C.",
""
],
[
"Ma",
"H.",
""
],
[
"Ma",
"H. L.",
""
],
[
"Ma",
"J. L.",
""
],
[
"Ma",
"L. L.",
""
],
[
"Ma",
"L. R.",
""
],
[
"Ma",
"Q. M.",
""
],
[
"Ma",
"R. Q.",
""
],
[
"Ma",
"R. Y.",
""
],
[
"Ma",
"T.",
""
],
[
"Ma",
"X. T.",
""
],
[
"Ma",
"X. Y.",
""
],
[
"Ma",
"Y. M.",
""
],
[
"Maas",
"F. E.",
""
],
[
"MacKay",
"I.",
""
],
[
"Maggiora",
"M.",
""
],
[
"Malde",
"S.",
""
],
[
"Mao",
"Y. J.",
""
],
[
"Mao",
"Z. P.",
""
],
[
"Marcello",
"S.",
""
],
[
"Meng",
"Y. H.",
""
],
[
"Meng",
"Z. X.",
""
],
[
"Messchendorp",
"J. G.",
""
],
[
"Mezzadri",
"G.",
""
],
[
"Miao",
"H.",
""
],
[
"Min",
"T. J.",
""
],
[
"Mitchell",
"R. E.",
""
],
[
"Mo",
"X. H.",
""
],
[
"Moses",
"B.",
""
],
[
"Muchnoi",
"N. Yu.",
""
],
[
"Muskalla",
"J.",
""
],
[
"Nefedov",
"Y.",
""
],
[
"Nerling",
"F.",
""
],
[
"Nie",
"L. S.",
""
],
[
"Nikolaev",
"I. B.",
""
],
[
"Ning",
"Z.",
""
],
[
"Nisar",
"S.",
""
],
[
"Niu",
"Q. L.",
""
],
[
"Niu",
"W. D.",
""
],
[
"Niu",
"Y.",
""
],
[
"Olsen",
"S. L.",
""
],
[
"Ouyang",
"Q.",
""
],
[
"Pacetti",
"S.",
""
],
[
"Pan",
"X.",
""
],
[
"Pan",
"Y.",
""
],
[
"Pathak",
"A.",
""
],
[
"Pei",
"Y. P.",
""
],
[
"Pelizaeus",
"M.",
""
],
[
"Peng",
"H. P.",
""
],
[
"Peng",
"Y. Y.",
""
],
[
"Peters",
"K.",
""
],
[
"Ping",
"J. L.",
""
],
[
"Ping",
"R. G.",
""
],
[
"Plura",
"S.",
""
],
[
"Prasad",
"V.",
""
],
[
"Qi",
"F. Z.",
""
],
[
"Qi",
"H. R.",
""
],
[
"Qi",
"M.",
""
],
[
"Qian",
"S.",
""
],
[
"Qian",
"W. B.",
""
],
[
"Qiao",
"C. F.",
""
],
[
"Qiao",
"J. H.",
""
],
[
"Qin",
"J. J.",
""
],
[
"Qin",
"L. Q.",
""
],
[
"Qin",
"L. Y.",
""
],
[
"Qin",
"X. P.",
""
],
[
"Qin",
"X. S.",
""
],
[
"Qin",
"Z. H.",
""
],
[
"Qiu",
"J. F.",
""
],
[
"Qu",
"Z. H.",
""
],
[
"Redmer",
"C. F.",
""
],
[
"Ren",
"K. J.",
""
],
[
"Rivetti",
"A.",
""
],
[
"Rolo",
"M.",
""
],
[
"Rong",
"G.",
""
],
[
"Rosner",
"Ch.",
""
],
[
"Ruan",
"M. Q.",
""
],
[
"Ruan",
"S. N.",
""
],
[
"Salone",
"N.",
""
],
[
"Sarantsev",
"A.",
""
],
[
"Schelhaas",
"Y.",
""
],
[
"Schoenning",
"K.",
""
],
[
"Scodeggio",
"M.",
""
],
[
"Shan",
"K. Y.",
""
],
[
"Shan",
"W.",
""
],
[
"Shan",
"X. Y.",
""
],
[
"Shang",
"Z. J.",
""
],
[
"Shangguan",
"J. F.",
""
],
[
"Shao",
"L. G.",
""
],
[
"Shao",
"M.",
""
],
[
"Shen",
"C. P.",
""
],
[
"Shen",
"H. F.",
""
],
[
"Shen",
"W. H.",
""
],
[
"Shen",
"X. Y.",
""
],
[
"Shi",
"B. A.",
""
],
[
"Shi",
"H.",
""
],
[
"Shi",
"J. L.",
""
],
[
"Shi",
"J. Y.",
""
],
[
"Shi",
"S. Y.",
""
],
[
"Shi",
"X.",
""
],
[
"Song",
"J. J.",
""
],
[
"Song",
"T. Z.",
""
],
[
"Song",
"W. M.",
""
],
[
"Song",
"Y. J.",
""
],
[
"Song",
"Y. X.",
""
],
[
"Sosio",
"S.",
""
],
[
"Spataro",
"S.",
""
],
[
"Stieler",
"F.",
""
],
[
"Su",
"S. S",
""
],
[
"Su",
"Y. J.",
""
],
[
"Sun",
"G. B.",
""
],
[
"Sun",
"G. X.",
""
],
[
"Sun",
"H.",
""
],
[
"Sun",
"H. K.",
""
],
[
"Sun",
"J. F.",
""
],
[
"Sun",
"K.",
""
],
[
"Sun",
"L.",
""
],
[
"Sun",
"S. S.",
""
],
[
"Sun",
"T.",
""
],
[
"Sun",
"Y. J.",
""
],
[
"Sun",
"Y. Z.",
""
],
[
"Sun",
"Z. Q.",
""
],
[
"Sun",
"Z. T.",
""
],
[
"Tang",
"C. J.",
""
],
[
"Tang",
"G. Y.",
""
],
[
"Tang",
"J.",
""
],
[
"Tang",
"M.",
""
],
[
"Tang",
"Y. A.",
""
],
[
"Tao",
"L. Y.",
""
],
[
"Tat",
"M.",
""
],
[
"Teng",
"J. X.",
""
],
[
"Thoren",
"V.",
""
],
[
"Tian",
"W. H.",
""
],
[
"Tian",
"Y.",
""
],
[
"Tian",
"Z. F.",
""
],
[
"Uman",
"I.",
""
],
[
"Wan",
"Y.",
""
],
[
"Wang",
"S. J.",
""
],
[
"Wang",
"B.",
""
],
[
"Wang",
"Bo",
""
],
[
"Wang",
"C.",
""
],
[
"Wang",
"D. Y.",
""
],
[
"Wang",
"H. J.",
""
],
[
"Wang",
"J. J.",
""
],
[
"Wang",
"J. P.",
""
],
[
"Wang",
"K.",
""
],
[
"Wang",
"L. L.",
""
],
[
"Wang",
"L. W.",
""
],
[
"Wang",
"M.",
""
],
[
"Wang",
"N. Y.",
""
],
[
"Wang",
"S.",
""
],
[
"Wang",
"S.",
""
],
[
"Wang",
"T.",
""
],
[
"Wang",
"T. J.",
""
],
[
"Wang",
"W.",
""
],
[
"Wang",
"W.",
""
],
[
"Wang",
"W. P.",
""
],
[
"Wang",
"X.",
""
],
[
"Wang",
"X. F.",
""
],
[
"Wang",
"X. J.",
""
],
[
"Wang",
"X. L.",
""
],
[
"Wang",
"X. N.",
""
],
[
"Wang",
"Y.",
""
],
[
"Wang",
"Y. D.",
""
],
[
"Wang",
"Y. F.",
""
],
[
"Wang",
"Y. H.",
""
],
[
"Wang",
"Y. L.",
""
],
[
"Wang",
"Y. N.",
""
],
[
"Wang",
"Y. Q.",
""
],
[
"Wang",
"Yaqian",
""
],
[
"Wang",
"Yi",
""
],
[
"Wang",
"Z.",
""
],
[
"Wang",
"Z. L.",
""
],
[
"Wang",
"Z. Y.",
""
],
[
"Wei",
"D. H.",
""
],
[
"Weidner",
"F.",
""
],
[
"Wen",
"S. P.",
""
],
[
"Wen",
"Y. R.",
""
],
[
"Wiedner",
"U.",
""
],
[
"Wilkinson",
"G.",
""
],
[
"Wolke",
"M.",
""
],
[
"Wollenberg",
"L.",
""
],
[
"Wu",
"C.",
""
],
[
"Wu",
"J. F.",
""
],
[
"Wu",
"L. H.",
""
],
[
"Wu",
"L. J.",
""
],
[
"Wu",
"Lianjie",
""
],
[
"Wu",
"X.",
""
],
[
"Wu",
"X. H.",
""
],
[
"Wu",
"Y. H.",
""
],
[
"Wu",
"Y. J.",
""
],
[
"Wu",
"Z.",
""
],
[
"Xia",
"L.",
""
],
[
"Xian",
"X. M.",
""
],
[
"Xiang",
"B. H.",
""
],
[
"Xiang",
"T.",
""
],
[
"Xiao",
"D.",
""
],
[
"Xiao",
"G. Y.",
""
],
[
"Xiao",
"H.",
""
],
[
"Xiao",
"Y. L.",
""
],
[
"Xiao",
"Z. J.",
""
],
[
"Xie",
"C.",
""
],
[
"Xie",
"X. H.",
""
],
[
"Xie",
"Y.",
""
],
[
"Xie",
"Y. G.",
""
],
[
"Xie",
"Y. H.",
""
],
[
"Xie",
"Z. P.",
""
],
[
"Xing",
"T. Y.",
""
],
[
"Xu",
"C. F.",
""
],
[
"Xu",
"C. J.",
""
],
[
"Xu",
"G. F.",
""
],
[
"Xu",
"M.",
""
],
[
"Xu",
"Q. J.",
""
],
[
"Xu",
"Q. N.",
""
],
[
"Xu",
"W. L.",
""
],
[
"Xu",
"X. P.",
""
],
[
"Xu",
"Y.",
""
],
[
"Xu",
"Y. C.",
""
],
[
"Xu",
"Z. S.",
""
],
[
"Yan",
"F.",
""
],
[
"Yan",
"L.",
""
],
[
"Yan",
"W. B.",
""
],
[
"Yan",
"W. C.",
""
],
[
"Yan",
"W. P.",
""
],
[
"Yan",
"X. Q.",
""
],
[
"Yang",
"H. J.",
""
],
[
"Yang",
"H. L.",
""
],
[
"Yang",
"H. X.",
""
],
[
"Yang",
"J. H.",
""
],
[
"Yang",
"R. J.",
""
],
[
"Yang",
"T.",
""
],
[
"Yang",
"Y.",
""
],
[
"Yang",
"Y. F.",
""
],
[
"Yang",
"Y. X.",
""
],
[
"Yang",
"Y. Z.",
""
],
[
"Yang",
"Z. W.",
""
],
[
"Yao",
"Z. P.",
""
],
[
"Ye",
"M.",
""
],
[
"Ye",
"M. H.",
""
],
[
"Yin",
"Junhao",
""
],
[
"You",
"Z. Y.",
""
],
[
"Yu",
"B. X.",
""
],
[
"Yu",
"C. X.",
""
],
[
"Yu",
"G.",
""
],
[
"Yu",
"J. S.",
""
],
[
"Yu",
"M. C.",
""
],
[
"Yu",
"T.",
""
],
[
"Yu",
"X. D.",
""
],
[
"Yuan",
"C. Z.",
""
],
[
"Yuan",
"J.",
""
],
[
"Yuan",
"J.",
""
],
[
"Yuan",
"L.",
""
],
[
"Yuan",
"S. C.",
""
],
[
"Yuan",
"Y.",
""
],
[
"Yuan",
"Z. Y.",
""
],
[
"Yue",
"C. X.",
""
],
[
"Yue",
"Ying",
""
],
[
"Zafar",
"A. A.",
""
],
[
"Zeng",
"F. R.",
""
],
[
"Zeng",
"S. H.",
""
],
[
"Zeng",
"X.",
""
],
[
"Zeng",
"Y.",
""
],
[
"Zeng",
"Y. J.",
""
],
[
"Zeng",
"Y. J.",
""
],
[
"Zhai",
"X. Y.",
""
],
[
"Zhai",
"Y. C.",
""
],
[
"Zhan",
"Y. H.",
""
],
[
"Zhang",
"A. Q.",
""
],
[
"Zhang",
"B. L.",
""
],
[
"Zhang",
"B. X.",
""
],
[
"Zhang",
"D. H.",
""
],
[
"Zhang",
"G. Y.",
""
],
[
"Zhang",
"H.",
""
],
[
"Zhang",
"H.",
""
],
[
"Zhang",
"H. C.",
""
],
[
"Zhang",
"H. H.",
""
],
[
"Zhang",
"H. Q.",
""
],
[
"Zhang",
"H. R.",
""
],
[
"Zhang",
"H. Y.",
""
],
[
"Zhang",
"J.",
""
],
[
"Zhang",
"J.",
""
],
[
"Zhang",
"J. J.",
""
],
[
"Zhang",
"J. L.",
""
],
[
"Zhang",
"J. Q.",
""
],
[
"Zhang",
"J. S.",
""
],
[
"Zhang",
"J. W.",
""
],
[
"Zhang",
"J. X.",
""
],
[
"Zhang",
"J. Y.",
""
],
[
"Zhang",
"J. Z.",
""
],
[
"Zhang",
"Jianyu",
""
],
[
"Zhang",
"L. M.",
""
],
[
"Zhang",
"Lei",
""
],
[
"Zhang",
"P.",
""
],
[
"Zhang",
"Q.",
""
],
[
"Zhang",
"Q. Y.",
""
],
[
"Zhang",
"R. Y.",
""
],
[
"Zhang",
"S. H.",
""
],
[
"Zhang",
"Shulei",
""
],
[
"Zhang",
"X. M.",
""
],
[
"Zhang",
"X. Y",
""
],
[
"Zhang",
"X. Y.",
""
],
[
"Zhang",
"Y.",
""
],
[
"Zhang",
"Y.",
""
],
[
"Zhang",
"Y. T.",
""
],
[
"Zhang",
"Y. H.",
""
],
[
"Zhang",
"Y. M.",
""
],
[
"Zhang",
"Yan",
""
],
[
"Zhang",
"Z. D.",
""
],
[
"Zhang",
"Z. H.",
""
],
[
"Zhang",
"Z. L.",
""
],
[
"Zhang",
"Z. X.",
""
],
[
"Zhang",
"Z. Y.",
""
],
[
"Zhang",
"Z. Y.",
""
],
[
"Zhang",
"Z. Z.",
""
],
[
"Zhang",
"Zh. Zh.",
""
],
[
"Zhao",
"G.",
""
],
[
"Zhao",
"J. Y.",
""
],
[
"Zhao",
"J. Z.",
""
],
[
"Zhao",
"L.",
""
],
[
"Zhao",
"Lei",
""
],
[
"Zhao",
"M. G.",
""
],
[
"Zhao",
"N.",
""
],
[
"Zhao",
"R. P.",
""
],
[
"Zhao",
"S. J.",
""
],
[
"Zhao",
"Y. B.",
""
],
[
"Zhao",
"Y. X.",
""
],
[
"Zhao",
"Z. G.",
""
],
[
"Zhemchugov",
"A.",
""
],
[
"Zheng",
"B.",
""
],
[
"Zheng",
"B. M.",
""
],
[
"Zheng",
"J. P.",
""
],
[
"Zheng",
"W. J.",
""
],
[
"Zheng",
"X. R.",
""
],
[
"Zheng",
"Y. H.",
""
],
[
"Zhong",
"B.",
""
],
[
"Zhong",
"X.",
""
],
[
"Zhou",
"H.",
""
],
[
"Zhou",
"J. Y.",
""
],
[
"Zhou",
"S.",
""
],
[
"Zhou",
"X.",
""
],
[
"Zhou",
"X. K.",
""
],
[
"Zhou",
"X. R.",
""
],
[
"Zhou",
"X. Y.",
""
],
[
"Zhou",
"Y. Z.",
""
],
[
"Zhou",
"Z. C.",
""
],
[
"Zhu",
"A. N.",
""
],
[
"Zhu",
"J.",
""
],
[
"Zhu",
"K.",
""
],
[
"Zhu",
"K. J.",
""
],
[
"Zhu",
"K. S.",
""
],
[
"Zhu",
"L.",
""
],
[
"Zhu",
"L. X.",
""
],
[
"Zhu",
"S. H.",
""
],
[
"Zhu",
"T. J.",
""
],
[
"Zhu",
"W. D.",
""
],
[
"Zhu",
"W. J.",
""
],
[
"Zhu",
"W. Z.",
""
],
[
"Zhu",
"Y. C.",
""
],
[
"Zhu",
"Z. A.",
""
],
[
"Zou",
"J. H.",
""
],
[
"Zu",
"J.",
""
]
] |
2502.07464 | Arthur Alexandre | Arthur Alexandre and Thomas Gu\'erin and David S. Dean | Effective description of Taylor dispersion in strongly corrugated channels | 16 pages, 3 figures | null | null | null | cond-mat.stat-mech physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Taylor dispersion in periodic but highly corrugated channels is studied. Exact analytical expressions for the long-time diffusion constant and drift along the channel are derived to next-to-leading order in the limit of small channel period. Using these results we show how an effective model for Taylor dispersion in tortuous porous media can be framed in terms of dispersion in a uniform channel with absorption/desorption at its surface, an effective slip length for the flow at the surface and an effective, universal, diffusion constant on the surface. This work thus extends the concept of an effective slip-length for hydrodynamics flows to Taylor dispersion by those flows. The analytical results are confirmed by numerical calculations, and present a robust method to understand and upscale the transport properties of flows in porous media. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 11:16:54 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 14:44:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Alexandre",
"Arthur",
""
],
[
"Guérin",
"Thomas",
""
],
[
"Dean",
"David S.",
""
]
] |
2502.07585 | Bassel Tarbush | Bary S.R. Pradelski, Bassel Tarbush | On the existence of pure epsilon-equilibrium | null | null | null | null | econ.TH cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that for any $\epsilon>0$, as the number of agents gets large, the share of games that admit a pure $\epsilon$-equilibrium converges to 1. Our result holds even for pure $\epsilon$-equilibrium in which all agents, except for at most one, play a best response. In contrast, it is known that the share of games that admit a pure Nash equilibrium, that is, for $\epsilon=0$, is asymptotically $1-1/e\approx 0.63$. This suggests that very small deviations from perfect rationality, captured by positive values of $\epsilon$, suffice to ensure the general existence of stable outcomes. We also study the existence of pure $\epsilon$-equilibrium when the number of actions gets large. Our proofs rely on the probabilistic method and on the Chen-Stein method. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 14:32:12 GMT"
},
{
"version": "v2",
"created": "Sun, 11 May 2025 14:46:15 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 09:43:05 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Pradelski",
"Bary S. R.",
""
],
[
"Tarbush",
"Bassel",
""
]
] |
2502.07613 | Emmanuele Battista Dr. | Kimet Jusufi, Abdelrahman Yasser, Emmanuele Battista, Nader Inan | Signatures of modified gravity from the gravitational Aharonov-Bohm effect | v2: few minor changes; accepted for publication in JCAP | null | 10.1088/1475-7516/2025/05/089 | null | gr-qc astro-ph.HE hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To date, no observational confirmation of dark matter particles has been found. In this paper, we put forward an alternative approach to inferring evidence for dark matter through modified gravity, without invoking fundamental dark matter particles. Specifically, we explore the possibility of extracting signatures of Kaluza-Klein gravity through the gravitational Aharonov-Bohm effect. Kaluza-Klein theory has recently been proposed as an alternative to the dark sector, and predicts a tower of particles, including spin-0 and spin-1 gravitons alongside the usual spin-2 gravitons, which can gravitationally couple to matter. We thus analyze a quantum system in free fall around a gravitating body in the presence of a modified Yukawa-like gravitational potential, and determine the gravitational phase induced by the additional degrees of freedom introduced by the Kaluza-Klein model. Our results reveal that, in addition to the usual result from General Relativity, the quantum wave function of the system exhibits an additional effect: a splitting of the energy levels with a new quantum number due to the extra vector gravitational degrees of freedom. The energy splitting difference between general relativity and Kaluza-Klein gravity is found to be of the order of meV for an atomic system and eV for a nuclear system. Similar values also arise in generic modified gravity models and can be feasibly tested in the future. Numerical estimates for the graviton mass are also provided, and potential imprints on gravitational waves are mentioned. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 15:04:34 GMT"
},
{
"version": "v2",
"created": "Fri, 16 May 2025 08:03:40 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Jusufi",
"Kimet",
""
],
[
"Yasser",
"Abdelrahman",
""
],
[
"Battista",
"Emmanuele",
""
],
[
"Inan",
"Nader",
""
]
] |
2502.07708 | Matthew Kvalheim | Matthew D. Kvalheim and Eduardo D. Sontag | Global linearization of asymptotically stable systems without hyperbolicity | 7 pages | null | null | null | math.DS cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We give a proof of an extension of the Hartman-Grobman theorem to nonhyperbolic but asymptotically stable equilibria of vector fields. Moreover, the linearizing topological conjugacy is (i) defined on the entire basin of attraction if the vector field is complete, and (ii) a $C^{k\geq 1}$-diffeomorphism on the complement of the equilibrium if the vector field is $C^k$ and the underlying space is not $5$-dimensional. We also show that the $C^k$ statement in the $5$-dimensional case is equivalent to the $4$-dimensional smooth Poincar\'{e} conjecture. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 17:05:07 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Feb 2025 17:07:43 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 17:45:55 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kvalheim",
"Matthew D.",
""
],
[
"Sontag",
"Eduardo D.",
""
]
] |
2502.07746 | Siddharth Viswanath | Siddharth Viswanath, Hiren Madhu, Dhananjay Bhaskar, Jake Kovalic, David R Johnson, Christopher Tape, Ian Adelstein, Rex Ying, Michael Perlmutter, Smita Krishnaswamy | HiPoNet: A Multi-View Simplicial Complex Network for High Dimensional Point-Cloud and Single-Cell Data | null | null | null | null | cs.LG math.AT | http://creativecommons.org/licenses/by/4.0/ | In this paper, we propose HiPoNet, an end-to-end differentiable neural network for regression, classification, and representation learning on high-dimensional point clouds. Our work is motivated by single-cell data which can have very high-dimensionality --exceeding the capabilities of existing methods for point clouds which are mostly tailored for 3D data. Moreover, modern single-cell and spatial experiments now yield entire cohorts of datasets (i.e., one data set for every patient), necessitating models that can process large, high-dimensional point-clouds at scale. Most current approaches build a single nearest-neighbor graph, discarding important geometric and topological information. In contrast, HiPoNet models the point-cloud as a set of higher-order simplicial complexes, with each particular complex being created using a reweighting of features. This method thus generates multiple constructs corresponding to different views of high-dimensional data, which in biology offers the possibility of disentangling distinct cellular processes. It then employs simplicial wavelet transforms to extract multiscale features, capturing both local and global topology from each view. We show that geometric and topological information is preserved in this framework both theoretically and empirically. We showcase the utility of HiPoNet on point-cloud level tasks, involving classification and regression of entire point-clouds in data cohorts. Experimentally, we find that HiPoNet outperforms other point-cloud and graph-based models on single-cell data. We also apply HiPoNet to spatial transcriptomics datasets using spatial coordinates as one of the views. Overall, HiPoNet offers a robust and scalable solution for high-dimensional data analysis. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 18:13:29 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 21:37:22 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Viswanath",
"Siddharth",
""
],
[
"Madhu",
"Hiren",
""
],
[
"Bhaskar",
"Dhananjay",
""
],
[
"Kovalic",
"Jake",
""
],
[
"Johnson",
"David R",
""
],
[
"Tape",
"Christopher",
""
],
[
"Adelstein",
"Ian",
""
],
[
"Ying",
"Rex",
""
],
[
"Perlmutter",
"Michael",
""
],
[
"Krishnaswamy",
"Smita",
""
]
] |
2502.07923 | Nikita Kornilov | Nikita Kornilov, Philip Zmushko, Andrei Semenov, Mark Ikonnikov, Alexander Gasnikov, Alexander Beznosikov | Sign Operator for Coping with Heavy-Tailed Noise in Non-Convex Optimization: High Probability Bounds Under $(L_0, L_1)$-Smoothness | null | null | null | null | math.OC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, non-convex optimization problems are more often described by generalized $(L_0, L_1)$-smoothness assumption rather than standard one. Meanwhile, severely corrupted data used in these problems has increased the demand for methods capable of handling heavy-tailed noises, i.e., noises with bounded $\kappa$-th moment. Motivated by these real-world trends and challenges, we explore sign-based methods in this setup and demonstrate their effectiveness in comparison with other popular solutions like clipping or normalization.
In theory, we prove the first-known high probability convergence bounds under $(L_0, L_1)$-smoothness and heavy-tailed noises with mild parameter dependencies. In the case of standard smoothness, these bounds are novel for sign-based methods as well. In particular, SignSGD with batching achieves sample complexity $\tilde{O}\left(\left(\frac{\Delta L_0d}{\varepsilon^2} + \frac{\Delta L_1d^\frac{3}{2}}{\varepsilon}\right)\left[1 + \left(\frac{\sigma}{\varepsilon}\right)^\frac{\kappa}{\kappa-1}\right]\right), \kappa \in (1,2]$. Under the assumption of symmetric noises, SignSGD with Majority Voting can robustly work on the whole range of $\kappa \in (0,2]$ with complexity $\tilde{O}\left(\left(\frac{\Delta L_0d}{\varepsilon^2} + \frac{\Delta L_1d^\frac{3}{2}}{\varepsilon}\right)\left[\frac{1}{\kappa^2} + \frac{\sigma^2}{\varepsilon^2}\right]\right)$. We also obtain results for parameter-agnostic setups, Polyak-Lojasiewicz functions and momentum-based methods (in expectation). Our theoretical findings are supported by the superior performance of sign-based methods in training Large Language Models compared to clipping and normalization. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 19:54:11 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:31:20 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kornilov",
"Nikita",
""
],
[
"Zmushko",
"Philip",
""
],
[
"Semenov",
"Andrei",
""
],
[
"Ikonnikov",
"Mark",
""
],
[
"Gasnikov",
"Alexander",
""
],
[
"Beznosikov",
"Alexander",
""
]
] |
2502.08005 | Wenlin Chen | Mingtian Zhang and Wenlin Chen and Jiajun He and Zijing Ou and Jos\'e Miguel Hern\'andez-Lobato and Bernhard Sch\"olkopf and David Barber | Towards Training One-Step Diffusion Models Without Distillation | 21 pages, 8 figures, 3 tables, 2 algorithms | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in training one-step diffusion models typically follow a two-stage pipeline: first training a teacher diffusion model and then distilling it into a one-step student model. This process often depends on both the teacher's score function for supervision and its weights for initializing the student model. In this paper, we explore whether one-step diffusion models can be trained directly without this distillation procedure. We introduce a family of new training methods that entirely forgo teacher score supervision, yet outperforms most teacher-guided distillation approaches. This suggests that score supervision is not essential for effective training of one-step diffusion models. However, we find that initializing the student model with the teacher's weights remains critical. Surprisingly, the key advantage of teacher initialization is not due to better latent-to-output mappings, but rather the rich set of feature representations across different noise levels that the teacher diffusion model provides. These insights take us one step closer towards training one-step diffusion models without distillation and provide a better understanding of the roles of teacher supervision and initialization in the distillation process. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 23:02:14 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Mar 2025 10:38:34 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 09:25:52 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhang",
"Mingtian",
""
],
[
"Chen",
"Wenlin",
""
],
[
"He",
"Jiajun",
""
],
[
"Ou",
"Zijing",
""
],
[
"Hernández-Lobato",
"José Miguel",
""
],
[
"Schölkopf",
"Bernhard",
""
],
[
"Barber",
"David",
""
]
] |
2502.08528 | Cuihong Wen | Ao liu, Zelin Zhang, Songbai Chen, Cuihong Wen and Jieci Wang | BCDDM: Branch-Corrected Denoising Diffusion Model for Black Hole Image Generation | 20 pages, 10figures | Astrophysical Journal Supplement Series, 2025 | null | null | astro-ph.GA cs.CV | http://creativecommons.org/licenses/by/4.0/ | The properties of black holes and accretion flows can be inferred by fitting Event Horizon Telescope (EHT) data to simulated images generated through general relativistic ray tracing (GRRT). However, due to the computationally intensive nature of GRRT, the efficiency of generating specific radiation flux images needs to be improved. This paper introduces the Branch Correction Denoising Diffusion Model (BCDDM), a deep learning framework that synthesizes black hole images directly from physical parameters. The model incorporates a branch correction mechanism and a weighted mixed loss function to enhance accuracy and stability. We have constructed a dataset of 2,157 GRRT-simulated images for training the BCDDM, which spans seven key physical parameters of the radiatively inefficient accretion flow (RIAF) model. Our experiments show a strong correlation between the generated images and their physical parameters. By enhancing the GRRT dataset with BCDDM-generated images and using ResNet50 for parameter regression, we achieve significant improvements in parameter prediction performance. BCDDM offers a novel approach to reducing the computational costs of black hole image generation, providing a faster and more efficient pathway for dataset augmentation, parameter estimation, and model fitting. | [
{
"version": "v1",
"created": "Wed, 12 Feb 2025 16:05:46 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Apr 2025 13:16:08 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:27:33 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"liu",
"Ao",
""
],
[
"Zhang",
"Zelin",
""
],
[
"Chen",
"Songbai",
""
],
[
"Wen",
"Cuihong",
""
],
[
"Wang",
"Jieci",
""
]
] |
2502.08736 | Wenlong Chen | Wenlong Chen, Naoki Kiyohara, Harrison Bo Hua Zhu, Jacob Curran-Sebastian, Samir Bhatt, Yingzhen Li | Recurrent Memory for Online Interdomain Gaussian Processes | 27 pages, 17 figures | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We propose a novel online Gaussian process (GP) model that is capable of capturing long-term memory in sequential data in an online learning setting. Our model, Online HiPPO Sparse Variational Gaussian Process (OHSVGP), leverages the HiPPO (High-order Polynomial Projection Operators) framework, which is popularized in the RNN domain due to its long-range memory modeling capabilities. We interpret the HiPPO time-varying orthogonal projections as inducing variables with time-dependent orthogonal polynomial basis functions, which allows the SVGP inducing variables to memorize the process history. We show that the HiPPO framework fits naturally into the interdomain GP framework and demonstrate that the kernel matrices can also be updated online in a recurrence form based on the ODE evolution of HiPPO. We evaluate OHSVGP with online prediction for 1D time series, continual learning in discriminative GP model for data with multidimensional inputs, and deep generative modeling with sparse Gaussian process variational autoencoder, showing that it outperforms existing online GP methods in terms of predictive performance, long-term memory preservation, and computational efficiency. | [
{
"version": "v1",
"created": "Wed, 12 Feb 2025 19:18:50 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Apr 2025 21:46:53 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 12:19:12 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chen",
"Wenlong",
""
],
[
"Kiyohara",
"Naoki",
""
],
[
"Zhu",
"Harrison Bo Hua",
""
],
[
"Curran-Sebastian",
"Jacob",
""
],
[
"Bhatt",
"Samir",
""
],
[
"Li",
"Yingzhen",
""
]
] |
2502.09008 | Julien Despres | Julien Despres, Leonardo Mazza, Marco Schir\`o | Dynamics of the Bose-Hubbard Model Induced by On-Site or Long-Range Two-Body Losses | 18 pages, 12 figures | null | null | null | cond-mat.quant-gas quant-ph | http://creativecommons.org/licenses/by/4.0/ | We present a theoretical study of the dissipative dynamics of the Bose-Hubbard model induced by on-site or long-range two-body losses. We first consider the one-dimensional chain and the two-dimensional square lattice, and study the dynamics induced by the sudden switch-on of two-body losses on a weakly-interacting superfluid state. The time-dependent density is obtained in the spirit of the Bogolyubov approach by calculating theoretically the equations of motion associated to the relevant quadratic bosonic correlators. In the one-dimensional case, our results compare very well with quasi-exact numerical calculations based on the quantum jump method implemented using tensor networks. We find that the intermediate-time dynamics of the density displays an algebraic decay characterized by an interaction-dependent power-law exponent. The latter property still holds for long-range two-body loss processes but it is absent in the two-dimensional square lattice with on-site losses. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 06:55:03 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 20:50:35 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Despres",
"Julien",
""
],
[
"Mazza",
"Leonardo",
""
],
[
"Schirò",
"Marco",
""
]
] |
2502.09192 | Myra Cheng | Lujain Ibrahim, Myra Cheng | Thinking beyond the anthropomorphic paradigm benefits LLM research | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anthropomorphism, or the attribution of human traits to technology, is an automatic and unconscious response that occurs even in those with advanced technical expertise. In this position paper, we analyze hundreds of thousands of research articles to present empirical evidence of the prevalence and growth of anthropomorphic terminology in research on large language models (LLMs). We argue for challenging the deeper assumptions reflected in this terminology -- which, though often useful, may inadvertently constrain LLM development -- and broadening beyond them to open new pathways for understanding and improving LLMs. Specifically, we identify and examine five anthropomorphic assumptions that shape research across the LLM development lifecycle. For each assumption (e.g., that LLMs must use natural language for reasoning, or that they should be evaluated on benchmarks originally meant for humans), we demonstrate empirical, non-anthropomorphic alternatives that remain under-explored yet offer promising directions for LLM research and development. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 11:32:09 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:24:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ibrahim",
"Lujain",
""
],
[
"Cheng",
"Myra",
""
]
] |
2502.09674 | Wenbo Pan | Wenbo Pan, Zhichao Liu, Qiguang Chen, Xiangyang Zhou, Haining Yu, and Xiaohua Jia | The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions | Code and artifacts: https://github.com/BMPixel/safety-residual-space Accepted by ICML 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models' safety-aligned behaviors, such as refusing harmful queries, can be represented by linear directions in activation space. Previous research modeled safety behavior with a single direction, limiting mechanistic understanding to an isolated safety feature. In this work, we discover that safety-aligned behavior is jointly controlled by multi-dimensional directions. Namely, we study the vector space of representation shifts during safety fine-tuning on Llama 3 8B for refusing jailbreaks. By studying orthogonal directions in the space, we first find that a dominant direction governs the model's refusal behavior, while multiple smaller directions represent distinct and interpretable features like hypothetical narrative and role-playing. We then measure how different directions promote or suppress the dominant direction, showing the important role of secondary directions in shaping the model's refusal representation. Finally, we demonstrate that removing certain trigger tokens in harmful queries can mitigate these directions to bypass the learned safety capability, providing new insights on understanding safety alignment vulnerability from a multi-dimensional perspective. Code and artifacts are available at https://github.com/BMPixel/safety-residual-space. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 06:39:22 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Feb 2025 03:24:45 GMT"
},
{
"version": "v3",
"created": "Thu, 22 May 2025 04:21:22 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 08:40:42 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Pan",
"Wenbo",
""
],
[
"Liu",
"Zhichao",
""
],
[
"Chen",
"Qiguang",
""
],
[
"Zhou",
"Xiangyang",
""
],
[
"Yu",
"Haining",
""
],
[
"Jia",
"Xiaohua",
""
]
] |
2502.09802 | Bharath Chowdhary Nagam | B. C. Nagam (1 and 2), J. A. Acevedo Barroso (3), J. Wilde (4), I. T. Andika (5 and 6), A. Manj\'on-Garc\'ia (7), R. Pearce-Casey (8), D. Stern (9), J. W. Nightingale (10), L. A. Moustakas (9), K. McCarthy (9), E. Moravec (11), L. Leuzzi (12 and 13), K. Rojas (14), S. Serjeant (8), T. E. Collett (15), P. Matavulj (14), M. Walmsley (16 and 17), B. Cl\'ement (3 and 18), C. Tortora (19), R. Gavazzi (20 and 21), R. B. Metcalf (12 and 13), C. M. O'Riordan (6), G. Verdoes Kleijn (1), L. V. E. Koopmans (1), E. A. Valentijn (1), V. Busillo (19 and 22 and 23), S. Schuldt (24 and 25), F. Courbin (4 and 26), G. Vernardos (27 and 28), M. Meneghetti (13 and 29), A. D\'iaz-S\'anchez (7), J. M. Diego (30), L. R. Ecker (31 and 32), T. T. Thai (20 and 33), A. R. Cooray (34), H. M. Courtois (35), L. Delchambre (36), G. Despali (12 and 13 and 29), D. Sluse (36), L. Ulivi (37 and 38 and 39), A. Melo (6 and 5), P. Corcho-Caballero (1), B. Altieri (40), A. Amara (41), S. Andreon (42), N. Auricchio (13), H. Aussel (43), C. Baccigalupi (44 and 45 and 46 and 47), M. Baldi (48 and 13 and 29), A. Balestra (49), S. Bardelli (13), P. Battaglia (13), D. Bonino (50), E. Branchini (51 and 52 and 42), M. Brescia (22 and 19), J. Brinchmann (53 and 54), A. Caillat (20), S. Camera (55 and 56 and 50), V. Capobianco (50), C. Carbone (25), J. Carretero (57 and 58), S. Casas (59), M. Castellano (60), G. Castignani (13), S. Cavuoti (19 and 23), A. Cimatti (61), C. Colodro-Conde (62), G. Congedo (63), C. J. Conselice (17), L. Conversi (64 and 40), Y. Copin (65), M. Cropper (66), A. Da Silva (67 and 68), H. Degaudenzi (69), G. De Lucia (45), A. M. Di Giorgio (70), J. Dinis (67 and 68), F. Dubath (69), C. A. J. Duncan (17), X. Dupac (40), S. Dusini (71), M. Fabricius (32 and 31), M. Farina (70), S. Farrens (43), S. Ferriol (65), M. Frailis (45), E. Franceschi (13), M. Fumana (25), K. George (31), W. Gillard (72), B. Gillis (63), C. Giocoli (13 and 29), P. G\'omez-Alvarez (73 and 40), A. Grazian (49), F. Grupp (32 and 31), L. Guzzo (24 and 42), S. V. H. Haugan (74), J. Hoar (40), W. Holmes (9), I. Hook (75), F. Hormuth (76), A. Hornstrup (77 and 78), P. Hudelot (21), K. Jahnke (79), M. Jhabvala (80), B. Joachimi (81), E. Keih\"anen (82), S. Kermiche (72), B. Kubik (65), K. Kuijken (83), M. K\"ummel (31), M. Kunz (84), H. Kurki-Suonio (85 and 86), R. Laureijs (87 and 1), D. Le Mignant (20), S. Ligori (50), P. B. Lilje (74), V. Lindholm (85 and 86), I. Lloro (88), G. Mainetti (89), E. Maiorano (13), O. Mansutti (45), O. Marggraf (90), K. Markovic (9), M. Martinelli (60 and 91), N. Martinet (20), F. Marulli (12 and 13 and 29), R. Massey (92), E. Medinaceli (13), M. Melchior (14), Y. Mellier (93 and 21), E. Merlin (60), G. Meylan (3), M. Moresco (12 and 13), L. Moscardini (12 and 13 and 29), R. Nakajima (90), C. Neissner (94 and 58), R. C. Nichol (41), S.-M. Niemi (87), C. Padilla (94), S. Paltani (69), F. Pasian (45), K. Pedersen (95), W. J. Percival (96 and 97 and 98), V. Pettorino (87), S. Pires (43), G. Polenta (99), M. Poncet (100), L. A. Popa (101), L. Pozzetti (13), F. Raison (32), R. Rebolo (62 and 102 and 103), A. Renzi (104 and 71), J. Rhodes (9), G. Riccio (19), E. Romelli (45), M. Roncarelli (13), E. Rossetti (48), R. Saglia (31 and 32), Z. Sakr (105 and 106 and 107), A. G. S\'anchez (32), D. Sapone (108), B. Sartoris (31 and 45), M. Schirmer (79), P. Schneider (90), T. Schrabback (109), A. Secroun (72), G. Seidel (79), S. Serrano (110 and 111 and 112), C. Sirignano (104 and 71), G. Sirri (29), J. Skottfelt (113), L. Stanco (71), J.-L. Starck (43), J. Steinwagner (32), P. Tallada-Cresp\'i (57 and 58), D. Tavagnacco (45), A. N. Taylor (63), H. I. Teplitz (114), I. Tereno (67 and 115), R. Toledo-Moreo (116), F. Torradeflot (58 and 57), A. Tsyganov (117), I. Tutusaus (106), L. Valenziano (13 and 118), T. Vassallo (31 and 45), A. Veropalumbo (42 and 52 and 51), Y. Wang (114), J. Weller (31 and 32), A. Zacchei (45 and 44), E. Zucca (13), C. Burigana (119 and 118), A. Mora (120), M. P\"ontinen (85), V. Scottez (93 and 121) ((1) Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands, (2) Minnesota Institute for Astrophysics, University of Minnesota, 116 Church St SE, Minneapolis, MN 55455, USA, (3) Institute of Physics, Laboratory of Astrophysics, Ecole Polytechnique F\'ed\'erale de Lausanne (EPFL), Observatoire de Sauverny, 1290 Versoix, Switzerland, (4) Institut de Ci\`encies del Cosmos (ICCUB), Universitat de Barcelona (IEEC-UB), Mart\'i i Franqu\`es 1, 08028 Barcelona, Spain, (5) Technical University of Munich, TUM School of Natural Sciences, Physics Department, James-Franck-Str.~1, 85748 Garching, Germany, (6) Max-Planck-Institut f\"ur Astrophysik, Karl-Schwarzschild-Str.~1, 85748 Garching, Germany, (7) Departamento F\'isica Aplicada, Universidad Polit\'ecnica de Cartagena, Campus Muralla del Mar, 30202 Cartagena, Murcia, Spain, (8) School of Physical Sciences, The Open University, Milton Keynes, MK7 6AA, UK, (9) Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA, (10) School of Mathematics, Statistics and Physics, Newcastle University, Herschel Building, Newcastle-upon-Tyne, NE1 7RU, UK, (11) Green Bank Observatory, P.O. Box 2, Green Bank, WV 24944, USA, (12) Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Universit\`a di Bologna, via Piero Gobetti 93/2, 40129 Bologna, Italy, (13) INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy, (14) University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland, (15) Institute of Cosmology and Gravitation, University of Portsmouth, Portsmouth PO1 3FX, UK, (16) David A. Dunlap Department of Astronomy \& Astrophysics, University of Toronto, 50 St George Street, Toronto, Ontario M5S 3H4, Canada, (17) Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, University of Manchester, Oxford Road, Manchester M13 9PL, UK, (18) SCITAS, Ecole Polytechnique F\'ed\'erale de Lausanne (EPFL), 1015 Lausanne, Switzerland, (19) INAF-Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy, (20) Aix-Marseille Universit\'e, CNRS, CNES, LAM, Marseille, France, (21) Institut d'Astrophysique de Paris, UMR 7095, CNRS, and Sorbonne Universit\'e, 98 bis boulevard Arago, 75014 Paris, France, (22) Department of Physics "E. Pancini", University Federico II, Via Cinthia 6, 80126, Napoli, Italy, (23) INFN section of Naples, Via Cinthia 6, 80126, Napoli, Italy, (24) Dipartimento di Fisica "Aldo Pontremoli", Universit\`a degli Studi di Milano, Via Celoria 16, 20133 Milano, Italy, (25) INAF-IASF Milano, Via Alfonso Corti 12, 20133 Milano, Italy, (26) Instituci\'o Catalana de Recerca i Estudis Avan\c{c}ats (ICREA), Passeig de Llu\'is Companys 23, 08010 Barcelona, Spain, (27) Department of Physics and Astronomy, Lehman College of the CUNY, Bronx, NY 10468, USA, (28) American Museum of Natural History, Department of Astrophysics, New York, NY 10024, USA, (29) INFN-Sezione di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy, (30) Instituto de F\'isica de Cantabria, Edificio Juan Jord\'a, Avenida de los Castros, 39005 Santander, Spain, (31) Universit\"ats-Sternwarte M\"unchen, Fakult\"at f\"ur Physik, Ludwig-Maximilians-Universit\"at M\"unchen, Scheinerstrasse 1, 81679 M\"unchen, Germany, (32) Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany, (33) National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan, (34) Department of Physics \& Astronomy, University of California Irvine, Irvine CA 92697, USA, (35) UCB Lyon 1, CNRS/IN2P3, IUF, IP2I Lyon, 4 rue Enrico Fermi, 69622 Villeurbanne, France, (36) STAR Institute, Quartier Agora - All\'ee du six Ao\^ut, 19c B-4000 Li\`ege, Belgium, (37) University of Trento, Via Sommarive 14, I-38123 Trento, Italy, (38) Dipartimento di Fisica e Astronomia, Universit\`a di Firenze, via G. Sansone 1, 50019 Sesto Fiorentino, Firenze, Italy, (39) INAF-Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125, Firenze, Italy, (40) ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Ca\~nada, Madrid, Spain, (41) School of Mathematics and Physics, University of Surrey, Guildford, Surrey, GU2 7XH, UK, (42) INAF-Osservatorio Astronomico di Brera, Via Brera 28, 20122 Milano, Italy, (43) Universit\'e Paris-Saclay, Universit\'e Paris Cit\'e, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France, (44) IFPU, Institute for Fundamental Physics of the Universe, via Beirut 2, 34151 Trieste, Italy, (45) INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy, (46) INFN, Sezione di Trieste, Via Valerio 2, 34127 Trieste TS, Italy, (47) SISSA, International School for Advanced Studies, Via Bonomea 265, 34136 Trieste TS, Italy, (48) Dipartimento di Fisica e Astronomia, Universit\`a di Bologna, Via Gobetti 93/2, 40129 Bologna, Italy, (49) INAF-Osservatorio Astronomico di Padova, Via dell'Osservatorio 5, 35122 Padova, Italy, (50) INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese (TO), Italy, (51) Dipartimento di Fisica, Universit\`a di Genova, Via Dodecaneso 33, 16146, Genova, Italy, (52) INFN-Sezione di Genova, Via Dodecaneso 33, 16146, Genova, Italy, (53) Instituto de Astrof\'isica e Ci\^encias do Espa\c{c}o, Universidade do Porto, CAUP, Rua das Estrelas, PT4150-762 Porto, Portugal, (54) Faculdade de Ci\^encias da Universidade do Porto, Rua do Campo de Alegre, 4150-007 Porto, Portugal, (55) Dipartimento di Fisica, Universit\`a degli Studi di Torino, Via P. Giuria 1, 10125 Torino, Italy, (56) INFN-Sezione di Torino, Via P. Giuria 1, 10125 Torino, Italy, (57) Centro de Investigaciones Energ\'eticas, Medioambientales y Tecnol\'ogicas (CIEMAT), Avenida Complutense 40, 28040 Madrid, Spain, (58) Port d'Informaci\'o Cient\'ifica, Campus UAB, C. Albareda s/n, 08193 Bellaterra (Barcelona), Spain, (59) Institute for Theoretical Particle Physics and Cosmology (TTK), RWTH Aachen University, 52056 Aachen, Germany, (60) INAF-Osservatorio Astronomico di Roma, Via Frascati 33, 00078 Monteporzio Catone, Italy, (61) Dipartimento di Fisica e Astronomia "Augusto Righi" - Alma Mater Studiorum Universit\`a di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy, (62) Instituto de Astrof\'isica de Canarias, V\'ia L\'actea, 38205 La Laguna, Tenerife, Spain, (63) Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK, (64) European Space Agency/ESRIN, Largo Galileo Galilei 1, 00044 Frascati, Roma, Italy, (65) Universit\'e Claude Bernard Lyon 1, CNRS/IN2P3, IP2I Lyon, UMR 5822, Villeurbanne, F-69100, France, (66) Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey RH5 6NT, UK, (67) Departamento de F\'isica, Faculdade de Ci\^encias, Universidade de Lisboa, Edif\'icio C8, Campo Grande, PT1749-016 Lisboa, Portugal, (68) Instituto de Astrof\'isica e Ci\^encias do Espa\c{c}o, Faculdade de Ci\^encias, Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal, (69) Department of Astronomy, University of Geneva, ch. d'Ecogia 16, 1290 Versoix, Switzerland, (70) INAF-Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere, 100, 00100 Roma, Italy, (71) INFN-Padova, Via Marzolo 8, 35131 Padova, Italy, (72) Aix-Marseille Universit\'e, CNRS/IN2P3, CPPM, Marseille, France, (73) FRACTAL S.L.N.E., calle Tulip\'an 2, Portal 13 1A, 28231, Las Rozas de Madrid, Spain, (74) Institute of Theoretical Astrophysics, University of Oslo, P.O. Box 1029 Blindern, 0315 Oslo, Norway, (75) Department of Physics, Lancaster University, Lancaster, LA1 4YB, UK, (76) Felix Hormuth Engineering, Goethestr. 17, 69181 Leimen, Germany, (77) Technical University of Denmark, Elektrovej 327, 2800 Kgs. Lyngby, Denmark, (78) Cosmic Dawn Center (DAWN), Denmark, (79) Max-Planck-Institut f\"ur Astronomie, K\"onigstuhl 17, 69117 Heidelberg, Germany, (80) NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA, (81) Department of Physics and Astronomy, University College London, Gower Street, London WC1E 6BT, UK, (82) Department of Physics and Helsinki Institute of Physics, Gustaf H\"allstr\"omin katu 2, 00014 University of Helsinki, Finland, (83) Leiden Observatory, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands, (84) Universit\'e de Gen\`eve, D\'epartement de Physique Th\'eorique and Centre for Astroparticle Physics, 24 quai Ernest-Ansermet, CH-1211 Gen\`eve 4, Switzerland, (85) Department of Physics, P.O. Box 64, 00014 University of Helsinki, Finland, (86) Helsinki Institute of Physics, Gustaf H\"allstr\"omin katu 2, University of Helsinki, Helsinki, Finland, (87) European Space Agency/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands, (88) NOVA optical infrared instrumentation group at ASTRON, Oude Hoogeveensedijk 4, 7991PD, Dwingeloo, The Netherlands, (89) Centre de Calcul de l'IN2P3/CNRS, 21 avenue Pierre de Coubertin 69627 Villeurbanne Cedex, France, (90) Universit\"at Bonn, Argelander-Institut f\"ur Astronomie, Auf dem H\"ugel 71, 53121 Bonn, Germany, (91) INFN-Sezione di Roma, Piazzale Aldo Moro, 2 - c/o Dipartimento di Fisica, Edificio G. Marconi, 00185 Roma, Italy, (92) Department of Physics, Institute for Computational Cosmology, Durham University, South Road, Durham, DH1 3LE, UK, (93) Institut d'Astrophysique de Paris, 98bis Boulevard Arago, 75014, Paris, France, (94) Institut de F\'isica d'Altes Energies (IFAE), The Barcelona Institute of Science and Technology, Campus UAB, 08193 Bellaterra (Barcelona), Spain, (95) DARK, Niels Bohr Institute, University of Copenhagen, Jagtvej 155, 2200 Copenhagen, Denmark, (96) Waterloo Centre for Astrophysics, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada, (97) Department of Physics and Astronomy, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada, (98) Perimeter Institute for Theoretical Physics, Waterloo, Ontario N2L 2Y5, Canada, (99) Space Science Data Center, Italian Space Agency, via del Politecnico snc, 00133 Roma, Italy, (100) Centre National d'Etudes Spatiales -- Centre spatial de Toulouse, 18 avenue Edouard Belin, 31401 Toulouse Cedex 9, France, (101) Institute of Space Science, Str. Atomistilor, nr. 409 M\u{a}gurele, Ilfov, 077125, Romania, (102) Consejo Superior de Investigaciones Cientificas, Calle Serrano 117, 28006 Madrid, Spain, (103) Universidad de La Laguna, Departamento de Astrof\'isica, 38206 La Laguna, Tenerife, Spain, (104) Dipartimento di Fisica e Astronomia "G. Galilei", Universit\`a di Padova, Via Marzolo 8, 35131 Padova, Italy, (105) Institut f\"ur Theoretische Physik, University of Heidelberg, Philosophenweg 16, 69120 Heidelberg, Germany, (106) Institut de Recherche en Astrophysique et Plan\'etologie (IRAP), Universit\'e de Toulouse, CNRS, UPS, CNES, 14 Av. Edouard Belin, 31400 Toulouse, France, (107) Universit\'e St Joseph; Faculty of Sciences, Beirut, Lebanon, (108) Departamento de F\'isica, FCFM, Universidad de Chile, Blanco Encalada 2008, Santiago, Chile, (109) Universit\"at Innsbruck, Institut f\"ur Astro- und Teilchenphysik, Technikerstr. 25/8, 6020 Innsbruck, Austria, (110) Institut d'Estudis Espacials de Catalunya (IEEC), Edifici RDIT, Campus UPC, 08860 Castelldefels, Barcelona, Spain, (111) Satlantis, University Science Park, Sede Bld 48940, Leioa-Bilbao, Spain, (112) Institute of Space Sciences (ICE, CSIC), Campus UAB, Carrer de Can Magrans, s/n, 08193 Barcelona, Spain, (113) Centre for Electronic Imaging, Open University, Walton Hall, Milton Keynes, MK7~6AA, UK, (114) Infrared Processing and Analysis Center, California Institute of Technology, Pasadena, CA 91125, USA, (115) Instituto de Astrof\'isica e Ci\^encias do Espa\c{c}o, Faculdade de Ci\^encias, Universidade de Lisboa, Tapada da Ajuda, 1349-018 Lisboa, Portugal, (116) Universidad Polit\'ecnica de Cartagena, Departamento de Electr\'onica y Tecnolog\'ia de Computadoras, Plaza del Hospital 1, 30202 Cartagena, Spain, (117) Centre for Information Technology, University of Groningen, P.O. Box 11044, 9700 CA Groningen, The Netherlands, (118) INFN-Bologna, Via Irnerio 46, 40126 Bologna, Italy, (119) INAF, Istituto di Radioastronomia, Via Piero Gobetti 101, 40129 Bologna, Italy, (120) Aurora Technology for European Space Agency (ESA), Camino bajo del Castillo, s/n, Urbanizacion Villafranca del Castillo, Villanueva de la Ca\~nada, 28692 Madrid, Spain, (121) ICL, Junia, Universit\'e Catholique de Lille, LITL, 59000 Lille, France) | Euclid: Finding strong gravitational lenses in the Early Release Observations using convolutional neural networks | null | null | null | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The Early Release Observations (ERO) from Euclid have detected several new galaxy-galaxy strong gravitational lenses, with the all-sky survey expected to find 170,000 new systems, greatly enhancing studies of dark matter, dark energy, and constraints on the cosmological parameters. As a first step, visual inspection of all galaxies in one of the ERO fields (Perseus) was carried out to identify candidate strong lensing systems and compared to the predictions from Convolutional Neural Networks (CNNs). However, the entire ERO data set is too large for expert visual inspection. In this paper, we therefore extend the CNN analysis to the whole ERO data set, using different CNN architectures and methodologies. Using five CNN architectures, we identified 8,469 strong gravitational lens candidates from IE-band cutouts of 13 Euclid ERO fields, narrowing them to 97 through visual inspection, including 14 grade A and 31 grade B candidates. We present the spectroscopic confirmation of a strong gravitational lensing candidate, EUCLJ081705.61+702348.8. The foreground lensing galaxy, an early-type system at redshift z = 0.335, and the background source, a star-forming galaxy at redshift z = 1.475 with [O II] emission, are both identified. Lens modeling using the Euclid strong lens modeling pipeline reveals two distinct arcs in a lensing configuration, with an Einstein radius of 1.18 \pm 0.03 arcseconds, confirming the lensing nature of the system. These findings highlight the importance of a broad CNN search to efficiently reduce candidates, followed by visual inspection to eliminate false positives and achieve a high-purity sample of strong lenses in Euclid. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 22:29:36 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Nagam",
"B. C.",
"",
"1 and 2"
],
[
"Barroso",
"J. A. Acevedo",
"",
"Institute of Physics, Laboratory of Astrophysics, Ecole Polytechnique Fédérale de Lausanne"
],
[
"Wilde",
"J.",
"",
"Institut de Ciències del Cosmos"
],
[
"Andika",
"I. T.",
"",
"5 and 6"
],
[
"Manjón-García",
"A.",
"",
"Departamento Física Aplicada, Universidad Politécnica de Cartagena, Campus Muralla del Mar, 30202 Cartagena, Murcia, Spain"
],
[
"Pearce-Casey",
"R.",
"",
"School of Physical Sciences, The Open University, Milton Keynes, MK7 6AA, UK"
],
[
"Stern",
"D.",
"",
"Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA"
],
[
"Nightingale",
"J. W.",
"",
"School of Mathematics, Statistics and Physics, Newcastle University, Herschel Building, Newcastle-upon-Tyne, NE1 7RU, UK"
],
[
"Moustakas",
"L. A.",
"",
"Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA"
],
[
"McCarthy",
"K.",
"",
"Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA"
],
[
"Moravec",
"E.",
"",
"Green Bank Observatory, P.O. Box 2, Green Bank, WV 24944, USA"
],
[
"Leuzzi",
"L.",
"",
"12 and 13"
],
[
"Rojas",
"K.",
"",
"University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland"
],
[
"Serjeant",
"S.",
"",
"School of Physical Sciences, The Open University, Milton Keynes, MK7 6AA, UK"
],
[
"Collett",
"T. E.",
"",
"Institute of Cosmology and Gravitation, University of Portsmouth, Portsmouth PO1 3FX, UK"
],
[
"Matavulj",
"P.",
"",
"University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland"
],
[
"Walmsley",
"M.",
"",
"16 and 17"
],
[
"Clément",
"B.",
"",
"3 and 18"
],
[
"Tortora",
"C.",
"",
"INAF-Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy"
],
[
"Gavazzi",
"R.",
"",
"20 and 21"
],
[
"Metcalf",
"R. B.",
"",
"12 and 13"
],
[
"O'Riordan",
"C. M.",
"",
"Max-Planck-Institut für Astrophysik, Karl-Schwarzschild-Str.~1, 85748 Garching, Germany"
],
[
"Kleijn",
"G. Verdoes",
"",
"Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands"
],
[
"Koopmans",
"L. V. E.",
"",
"Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands"
],
[
"Valentijn",
"E. A.",
"",
"Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands"
],
[
"Busillo",
"V.",
"",
"19 and 22 and 23"
],
[
"Schuldt",
"S.",
"",
"24 and 25"
],
[
"Courbin",
"F.",
"",
"4 and 26"
],
[
"Vernardos",
"G.",
"",
"27 and 28"
],
[
"Meneghetti",
"M.",
"",
"13 and 29"
],
[
"Díaz-Sánchez",
"A.",
"",
"Departamento Física Aplicada, Universidad Politécnica de Cartagena, Campus Muralla del Mar, 30202 Cartagena, Murcia, Spain"
],
[
"Diego",
"J. M.",
"",
"Instituto de Física de Cantabria, Edificio Juan Jordá, Avenida de los Castros, 39005 Santander, Spain"
],
[
"Ecker",
"L. R.",
"",
"31 and 32"
],
[
"Thai",
"T. T.",
"",
"20 and 33"
],
[
"Cooray",
"A. R.",
"",
"Department of Physics \\& Astronomy, University of California Irvine, Irvine CA 92697, USA"
],
[
"Courtois",
"H. M.",
"",
"UCB Lyon 1, CNRS/IN2P3, IUF, IP2I Lyon, 4 rue Enrico Fermi, 69622 Villeurbanne, France"
],
[
"Delchambre",
"L.",
"",
"STAR Institute, Quartier Agora - Allée du six Août, 19c B-4000 Liège, Belgium"
],
[
"Despali",
"G.",
"",
"12 and 13 and 29"
],
[
"Sluse",
"D.",
"",
"STAR Institute, Quartier Agora - Allée du six Août, 19c B-4000 Liège, Belgium"
],
[
"Ulivi",
"L.",
"",
"37 and 38 and 39"
],
[
"Melo",
"A.",
"",
"6 and 5"
],
[
"Corcho-Caballero",
"P.",
"",
"Kapteyn Astronomical Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands"
],
[
"Altieri",
"B.",
"",
"ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Cañada, Madrid, Spain"
],
[
"Amara",
"A.",
"",
"School of Mathematics and Physics, University of Surrey, Guildford, Surrey, GU2 7XH, UK"
],
[
"Andreon",
"S.",
"",
"INAF-Osservatorio Astronomico di Brera, Via Brera 28, 20122 Milano, Italy"
],
[
"Auricchio",
"N.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Aussel",
"H.",
"",
"Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France"
],
[
"Baccigalupi",
"C.",
"",
"44 and 45 and 46 and 47"
],
[
"Baldi",
"M.",
"",
"48 and 13 and 29"
],
[
"Balestra",
"A.",
"",
"INAF-Osservatorio Astronomico di Padova, Via dell'Osservatorio 5, 35122 Padova, Italy"
],
[
"Bardelli",
"S.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Battaglia",
"P.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Bonino",
"D.",
"",
"INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese"
],
[
"Branchini",
"E.",
"",
"51 and 52 and 42"
],
[
"Brescia",
"M.",
"",
"22 and 19"
],
[
"Brinchmann",
"J.",
"",
"53 and 54"
],
[
"Caillat",
"A.",
"",
"Aix-Marseille Université, CNRS, CNES, LAM, Marseille, France"
],
[
"Camera",
"S.",
"",
"55 and 56 and 50"
],
[
"Capobianco",
"V.",
"",
"INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese"
],
[
"Carbone",
"C.",
"",
"INAF-IASF Milano, Via Alfonso Corti 12, 20133 Milano, Italy"
],
[
"Carretero",
"J.",
"",
"57 and 58"
],
[
"Casas",
"S.",
"",
"Institute for Theoretical Particle Physics and Cosmology"
],
[
"Castellano",
"M.",
"",
"INAF-Osservatorio Astronomico di Roma, Via Frascati 33, 00078 Monteporzio Catone, Italy"
],
[
"Castignani",
"G.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Cavuoti",
"S.",
"",
"19 and 23"
],
[
"Cimatti",
"A.",
"",
"Dipartimento di Fisica e Astronomia \"Augusto Righi\" - Alma Mater Studiorum Università di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy"
],
[
"Colodro-Conde",
"C.",
"",
"Instituto de Astrofísica de Canarias, Vía Láctea, 38205 La Laguna, Tenerife, Spain"
],
[
"Congedo",
"G.",
"",
"Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK"
],
[
"Conselice",
"C. J.",
"",
"Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, University of Manchester, Oxford Road, Manchester M13 9PL, UK"
],
[
"Conversi",
"L.",
"",
"64 and 40"
],
[
"Copin",
"Y.",
"",
"Université Claude Bernard Lyon 1, CNRS/IN2P3, IP2I Lyon, UMR 5822, Villeurbanne, F-69100, France"
],
[
"Cropper",
"M.",
"",
"Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey RH5 6NT, UK"
],
[
"Da Silva",
"A.",
"",
"67 and 68"
],
[
"Degaudenzi",
"H.",
"",
"Department of Astronomy, University of Geneva, ch. d'Ecogia 16, 1290 Versoix, Switzerland"
],
[
"De Lucia",
"G.",
"",
"INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy"
],
[
"Di Giorgio",
"A. M.",
"",
"INAF-Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere, 100, 00100 Roma, Italy"
],
[
"Dinis",
"J.",
"",
"67 and 68"
],
[
"Dubath",
"F.",
"",
"Department of Astronomy, University of Geneva, ch. d'Ecogia 16, 1290 Versoix, Switzerland"
],
[
"Duncan",
"C. A. J.",
"",
"Jodrell Bank Centre for Astrophysics, Department of Physics and Astronomy, University of Manchester, Oxford Road, Manchester M13 9PL, UK"
],
[
"Dupac",
"X.",
"",
"ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Cañada, Madrid, Spain"
],
[
"Dusini",
"S.",
"",
"INFN-Padova, Via Marzolo 8, 35131 Padova, Italy"
],
[
"Fabricius",
"M.",
"",
"32 and 31"
],
[
"Farina",
"M.",
"",
"INAF-Istituto di Astrofisica e Planetologia Spaziali, via del Fosso del Cavaliere, 100, 00100 Roma, Italy"
],
[
"Farrens",
"S.",
"",
"Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France"
],
[
"Ferriol",
"S.",
"",
"Université Claude Bernard Lyon 1, CNRS/IN2P3, IP2I Lyon, UMR 5822, Villeurbanne, F-69100, France"
],
[
"Frailis",
"M.",
"",
"INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy"
],
[
"Franceschi",
"E.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Fumana",
"M.",
"",
"INAF-IASF Milano, Via Alfonso Corti 12, 20133 Milano, Italy"
],
[
"George",
"K.",
"",
"Universitäts-Sternwarte München, Fakultät für Physik, Ludwig-Maximilians-Universität München, Scheinerstrasse 1, 81679 München, Germany"
],
[
"Gillard",
"W.",
"",
"Aix-Marseille Université, CNRS/IN2P3, CPPM, Marseille, France"
],
[
"Gillis",
"B.",
"",
"Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK"
],
[
"Giocoli",
"C.",
"",
"13 and 29"
],
[
"Gómez-Alvarez",
"P.",
"",
"73 and 40"
],
[
"Grazian",
"A.",
"",
"INAF-Osservatorio Astronomico di Padova, Via dell'Osservatorio 5, 35122 Padova, Italy"
],
[
"Grupp",
"F.",
"",
"32 and 31"
],
[
"Guzzo",
"L.",
"",
"24 and 42"
],
[
"Haugan",
"S. V. H.",
"",
"Institute of Theoretical Astrophysics, University of Oslo, P.O. Box 1029 Blindern, 0315 Oslo, Norway"
],
[
"Hoar",
"J.",
"",
"ESAC/ESA, Camino Bajo del Castillo, s/n., Urb. Villafranca del Castillo, 28692 Villanueva de la Cañada, Madrid, Spain"
],
[
"Holmes",
"W.",
"",
"Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA"
],
[
"Hook",
"I.",
"",
"Department of Physics, Lancaster University, Lancaster, LA1 4YB, UK"
],
[
"Hormuth",
"F.",
"",
"Felix Hormuth Engineering, Goethestr. 17, 69181 Leimen, Germany"
],
[
"Hornstrup",
"A.",
"",
"77 and 78"
],
[
"Hudelot",
"P.",
"",
"Institut d'Astrophysique de Paris, UMR 7095, CNRS, and Sorbonne Université, 98 bis boulevard Arago, 75014 Paris, France"
],
[
"Jahnke",
"K.",
"",
"Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany"
],
[
"Jhabvala",
"M.",
"",
"NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA"
],
[
"Joachimi",
"B.",
"",
"Department of Physics and Astronomy, University College London, Gower Street, London WC1E 6BT, UK"
],
[
"Keihänen",
"E.",
"",
"Department of Physics and Helsinki Institute of Physics, Gustaf Hällströmin katu 2, 00014 University of Helsinki, Finland"
],
[
"Kermiche",
"S.",
"",
"Aix-Marseille Université, CNRS/IN2P3, CPPM, Marseille, France"
],
[
"Kubik",
"B.",
"",
"Université Claude Bernard Lyon 1, CNRS/IN2P3, IP2I Lyon, UMR 5822, Villeurbanne, F-69100, France"
],
[
"Kuijken",
"K.",
"",
"Leiden Observatory, Leiden University, Einsteinweg 55, 2333 CC Leiden, The Netherlands"
],
[
"Kümmel",
"M.",
"",
"Universitäts-Sternwarte München, Fakultät für Physik, Ludwig-Maximilians-Universität München, Scheinerstrasse 1, 81679 München, Germany"
],
[
"Kunz",
"M.",
"",
"Université de Genève, Département de Physique Théorique and Centre for Astroparticle Physics, 24 quai Ernest-Ansermet, CH-1211 Genève 4, Switzerland"
],
[
"Kurki-Suonio",
"H.",
"",
"85 and 86"
],
[
"Laureijs",
"R.",
"",
"87 and 1"
],
[
"Mignant",
"D. Le",
"",
"Aix-Marseille Université, CNRS, CNES, LAM, Marseille, France"
],
[
"Ligori",
"S.",
"",
"INAF-Osservatorio Astrofisico di Torino, Via Osservatorio 20, 10025 Pino Torinese"
],
[
"Lilje",
"P. B.",
"",
"Institute of Theoretical Astrophysics, University of Oslo, P.O. Box 1029 Blindern, 0315 Oslo, Norway"
],
[
"Lindholm",
"V.",
"",
"85 and 86"
],
[
"Lloro",
"I.",
"",
"NOVA optical infrared instrumentation group at ASTRON, Oude Hoogeveensedijk 4, 7991PD, Dwingeloo, The Netherlands"
],
[
"Mainetti",
"G.",
"",
"Centre de Calcul de l'IN2P3/CNRS, 21 avenue Pierre de Coubertin 69627 Villeurbanne Cedex, France"
],
[
"Maiorano",
"E.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Mansutti",
"O.",
"",
"INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy"
],
[
"Marggraf",
"O.",
"",
"Universität Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany"
],
[
"Markovic",
"K.",
"",
"Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA"
],
[
"Martinelli",
"M.",
"",
"60 and 91"
],
[
"Martinet",
"N.",
"",
"Aix-Marseille Université, CNRS, CNES, LAM, Marseille, France"
],
[
"Marulli",
"F.",
"",
"12 and 13 and 29"
],
[
"Massey",
"R.",
"",
"Department of Physics, Institute for Computational Cosmology, Durham University, South Road, Durham, DH1 3LE, UK"
],
[
"Medinaceli",
"E.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Melchior",
"M.",
"",
"University of Applied Sciences and Arts of Northwestern Switzerland, School of Engineering, 5210 Windisch, Switzerland"
],
[
"Mellier",
"Y.",
"",
"93 and 21"
],
[
"Merlin",
"E.",
"",
"INAF-Osservatorio Astronomico di Roma, Via Frascati 33, 00078 Monteporzio Catone, Italy"
],
[
"Meylan",
"G.",
"",
"Institute of Physics, Laboratory of Astrophysics, Ecole Polytechnique Fédérale de Lausanne"
],
[
"Moresco",
"M.",
"",
"12 and 13"
],
[
"Moscardini",
"L.",
"",
"12 and 13 and 29"
],
[
"Nakajima",
"R.",
"",
"Universität Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany"
],
[
"Neissner",
"C.",
"",
"94 and 58"
],
[
"Nichol",
"R. C.",
"",
"School of Mathematics and Physics, University of Surrey, Guildford, Surrey, GU2 7XH, UK"
],
[
"Niemi",
"S. -M.",
"",
"European Space Agency/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands"
],
[
"Padilla",
"C.",
"",
"Institut de Física d'Altes Energies"
],
[
"Paltani",
"S.",
"",
"Department of Astronomy, University of Geneva, ch. d'Ecogia 16, 1290 Versoix, Switzerland"
],
[
"Pasian",
"F.",
"",
"INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy"
],
[
"Pedersen",
"K.",
"",
"DARK, Niels Bohr Institute, University of Copenhagen, Jagtvej 155, 2200 Copenhagen, Denmark"
],
[
"Percival",
"W. J.",
"",
"96 and 97 and 98"
],
[
"Pettorino",
"V.",
"",
"European Space Agency/ESTEC, Keplerlaan 1, 2201 AZ Noordwijk, The Netherlands"
],
[
"Pires",
"S.",
"",
"Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France"
],
[
"Polenta",
"G.",
"",
"Space Science Data Center, Italian Space Agency, via del Politecnico snc, 00133 Roma, Italy"
],
[
"Poncet",
"M.",
"",
"Centre National d'Etudes Spatiales -- Centre spatial de Toulouse, 18 avenue Edouard Belin, 31401 Toulouse Cedex 9, France"
],
[
"Popa",
"L. A.",
"",
"Institute of Space Science, Str. Atomistilor, nr. 409 Măgurele, Ilfov, 077125, Romania"
],
[
"Pozzetti",
"L.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Raison",
"F.",
"",
"Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany"
],
[
"Rebolo",
"R.",
"",
"62 and 102 and 103"
],
[
"Renzi",
"A.",
"",
"104 and 71"
],
[
"Rhodes",
"J.",
"",
"Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109, USA"
],
[
"Riccio",
"G.",
"",
"INAF-Osservatorio Astronomico di Capodimonte, Via Moiariello 16, 80131 Napoli, Italy"
],
[
"Romelli",
"E.",
"",
"INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy"
],
[
"Roncarelli",
"M.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Rossetti",
"E.",
"",
"Dipartimento di Fisica e Astronomia, Università di Bologna, Via Gobetti 93/2, 40129 Bologna, Italy"
],
[
"Saglia",
"R.",
"",
"31 and 32"
],
[
"Sakr",
"Z.",
"",
"105 and 106 and 107"
],
[
"Sánchez",
"A. G.",
"",
"Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany"
],
[
"Sapone",
"D.",
"",
"Departamento de Física, FCFM, Universidad de Chile, Blanco Encalada 2008, Santiago, Chile"
],
[
"Sartoris",
"B.",
"",
"31 and 45"
],
[
"Schirmer",
"M.",
"",
"Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany"
],
[
"Schneider",
"P.",
"",
"Universität Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany"
],
[
"Schrabback",
"T.",
"",
"Universität Innsbruck, Institut für Astro- und Teilchenphysik, Technikerstr. 25/8, 6020 Innsbruck, Austria"
],
[
"Secroun",
"A.",
"",
"Aix-Marseille Université, CNRS/IN2P3, CPPM, Marseille, France"
],
[
"Seidel",
"G.",
"",
"Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany"
],
[
"Serrano",
"S.",
"",
"110 and 111 and 112"
],
[
"Sirignano",
"C.",
"",
"104 and 71"
],
[
"Sirri",
"G.",
"",
"INFN-Sezione di Bologna, Viale Berti Pichat 6/2, 40127 Bologna, Italy"
],
[
"Skottfelt",
"J.",
"",
"Centre for Electronic Imaging, Open University, Walton Hall, Milton Keynes, MK7~6AA, UK"
],
[
"Stanco",
"L.",
"",
"INFN-Padova, Via Marzolo 8, 35131 Padova, Italy"
],
[
"Starck",
"J. -L.",
"",
"Université Paris-Saclay, Université Paris Cité, CEA, CNRS, AIM, 91191, Gif-sur-Yvette, France"
],
[
"Steinwagner",
"J.",
"",
"Max Planck Institute for Extraterrestrial Physics, Giessenbachstr. 1, 85748 Garching, Germany"
],
[
"Tallada-Crespí",
"P.",
"",
"57 and 58"
],
[
"Tavagnacco",
"D.",
"",
"INAF-Osservatorio Astronomico di Trieste, Via G. B. Tiepolo 11, 34143 Trieste, Italy"
],
[
"Taylor",
"A. N.",
"",
"Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK"
],
[
"Teplitz",
"H. I.",
"",
"Infrared Processing and Analysis Center, California Institute of Technology, Pasadena, CA 91125, USA"
],
[
"Tereno",
"I.",
"",
"67 and 115"
],
[
"Toledo-Moreo",
"R.",
"",
"Universidad Politécnica de Cartagena, Departamento de Electrónica y Tecnología de Computadoras, Plaza del Hospital 1, 30202 Cartagena, Spain"
],
[
"Torradeflot",
"F.",
"",
"58 and 57"
],
[
"Tsyganov",
"A.",
"",
"Centre for Information Technology, University of Groningen, P.O. Box 11044, 9700 CA Groningen, The Netherlands"
],
[
"Tutusaus",
"I.",
"",
"Institut de Recherche en Astrophysique et Planétologie"
],
[
"Valenziano",
"L.",
"",
"13 and 118"
],
[
"Vassallo",
"T.",
"",
"31 and 45"
],
[
"Veropalumbo",
"A.",
"",
"42 and 52 and 51"
],
[
"Wang",
"Y.",
"",
"Infrared Processing and Analysis Center, California Institute of Technology, Pasadena, CA 91125, USA"
],
[
"Weller",
"J.",
"",
"31 and 32"
],
[
"Zacchei",
"A.",
"",
"45 and 44"
],
[
"Zucca",
"E.",
"",
"INAF-Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, Via Piero Gobetti 93/3, 40129 Bologna, Italy"
],
[
"Burigana",
"C.",
"",
"119 and 118"
],
[
"Mora",
"A.",
"",
"Aurora Technology for European Space Agency"
],
[
"Pöntinen",
"M.",
"",
"Department of Physics, P.O. Box 64, 00014 University of Helsinki, Finland"
],
[
"Scottez",
"V.",
"",
"93 and 121"
]
] |
2502.09908 | Haruki Watanabe | Rintaro Masaoka, Tomohiro Soejima, Haruki Watanabe | Rigorous lower bound of the dynamical critical exponent of the Ising model | 5+6 pages, v2: published version | J. Stat. Phys. 192, 76 (2025) | 10.1007/s10955-025-03456-3 | null | cond-mat.stat-mech cond-mat.other math-ph math.MP quant-ph | http://creativecommons.org/licenses/by/4.0/ | We study the kinetic Ising model under Glauber dynamics and establish an upper bound on the spectral gap for finite systems. This bound implies the critical exponent inequality $z \geq 2$, thereby rigorously improving the previously known estimate $z \geq 2 - \eta$. Our proof relies on the mapping from stochastic processes to frustration-free quantum systems and leverages the Simon--Lieb and Gosset--Huang inequalities. | [
{
"version": "v1",
"created": "Fri, 14 Feb 2025 04:40:24 GMT"
},
{
"version": "v2",
"created": "Wed, 21 May 2025 19:26:11 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Masaoka",
"Rintaro",
""
],
[
"Soejima",
"Tomohiro",
""
],
[
"Watanabe",
"Haruki",
""
]
] |
2502.10020 | Joongkyu Lee | Joongkyu Lee, Min-hwan Oh | Improved Online Confidence Bounds for Multinomial Logistic Bandits | Accepted at ICML 2025 | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we propose an improved online confidence bound for multinomial logistic (MNL) models and apply this result to MNL bandits, achieving variance-dependent optimal regret. Recently, Lee & Oh (2024) established an online confidence bound for MNL models and achieved nearly minimax-optimal regret in MNL bandits. However, their results still depend on the norm-boundedness of the unknown parameter $B$ and the maximum size of possible outcomes $K$. To address this, we first derive an online confidence bound of $O\left(\sqrt{d \log t} + B \right)$, which is a significant improvement over the previous bound of $O (B \sqrt{d} \log t \log K )$ (Lee & Oh, 2024). This is mainly achieved by establishing tighter self-concordant properties of the MNL loss and introducing a novel intermediary term to bound the estimation error. Using this new online confidence bound, we propose a constant-time algorithm, OFU-MNL++, which achieves a variance-dependent regret bound of $O \Big( d \log T \sqrt{ \sum_{t=1}^T \sigma_t^2 } \Big) $ for sufficiently large $T$, where $\sigma_t^2$ denotes the variance of the rewards at round $t$, $d$ is the dimension of the contexts, and $T$ is the total number of rounds. Furthermore, we introduce a Maximum Likelihood Estimation (MLE)-based algorithm, OFU-MN$^2$L, which achieves an anytime poly(B)-free regret of $O \Big( d \log (BT) \sqrt{ \sum_{t=1}^T \sigma_t^2 } \Big) $. | [
{
"version": "v1",
"created": "Fri, 14 Feb 2025 09:01:12 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Feb 2025 13:48:42 GMT"
},
{
"version": "v3",
"created": "Fri, 07 Mar 2025 04:29:35 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 13:28:48 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Lee",
"Joongkyu",
""
],
[
"Oh",
"Min-hwan",
""
]
] |
2502.10328 | Jiajun He | Leo Zhang, Peter Potaptchik, Jiajun He, Yuanqi Du, Arnaud Doucet, Francisco Vargas, Hai-Dang Dau, Saifuddin Syed | Accelerated Parallel Tempering via Neural Transports | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Markov Chain Monte Carlo (MCMC) algorithms are essential tools in computational statistics for sampling from unnormalised probability distributions, but can be fragile when targeting high-dimensional, multimodal, or complex target distributions. Parallel Tempering (PT) enhances MCMC's sample efficiency through annealing and parallel computation, propagating samples from tractable reference distributions to intractable targets via state swapping across interpolating distributions. The effectiveness of PT is limited by the often minimal overlap between adjacent distributions in challenging problems, which requires increasing the computational resources to compensate. We introduce a framework that accelerates PT by leveraging neural samplers-including normalising flows, diffusion models, and controlled diffusions-to reduce the required overlap. Our approach utilises neural samplers in parallel, circumventing the computational burden of neural samplers while preserving the asymptotic consistency of classical PT. We demonstrate theoretically and empirically on a variety of multimodal sampling problems that our method improves sample quality, reduces the computational cost compared to classical PT, and enables efficient free energies/normalising constants estimation. | [
{
"version": "v1",
"created": "Fri, 14 Feb 2025 17:41:44 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 14:46:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhang",
"Leo",
""
],
[
"Potaptchik",
"Peter",
""
],
[
"He",
"Jiajun",
""
],
[
"Du",
"Yuanqi",
""
],
[
"Doucet",
"Arnaud",
""
],
[
"Vargas",
"Francisco",
""
],
[
"Dau",
"Hai-Dang",
""
],
[
"Syed",
"Saifuddin",
""
]
] |
2502.10330 | Shutong Ding | Shutong Ding, Yimiao Zhou, Ke Hu, Xi Yao, Junchi Yan, Xiaoying Tang, Ye Shi | Exploring the Boundary of Diffusion-based Methods for Solving Constrained Optimization | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion models have achieved remarkable success in generative tasks such as image and video synthesis, and in control domains like robotics, owing to their strong generalization capabilities and proficiency in fitting complex multimodal distributions. However, their full potential in solving Continuous Constrained Optimization problems remains largely underexplored. Our work commences by investigating a two-dimensional constrained quadratic optimization problem as an illustrative example to explore the inherent challenges and issues when applying diffusion models to such optimization tasks and providing theoretical analyses for these observations. To address the identified gaps and harness diffusion models for Continuous Constrained Optimization, we build upon this analysis to propose a novel diffusion-based framework for optimization problems called DiOpt. This framework operates in two distinct phases: an initial warm-start phase, implemented via supervised learning, followed by a bootstrapping phase. This dual-phase architecture is designed to iteratively refine solutions, thereby improving the objective function while rigorously satisfying problem constraints. Finally, multiple candidate solutions are sampled, and the optimal one is selected through a screening process. We present extensive experiments detailing the training dynamics of DiOpt, its performance across a diverse set of Continuous Constrained Optimization problems, and an analysis of the impact of DiOpt's various hyperparameters. | [
{
"version": "v1",
"created": "Fri, 14 Feb 2025 17:43:08 GMT"
},
{
"version": "v2",
"created": "Sat, 24 May 2025 19:48:10 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 09:05:01 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ding",
"Shutong",
""
],
[
"Zhou",
"Yimiao",
""
],
[
"Hu",
"Ke",
""
],
[
"Yao",
"Xi",
""
],
[
"Yan",
"Junchi",
""
],
[
"Tang",
"Xiaoying",
""
],
[
"Shi",
"Ye",
""
]
] |
2502.10500 | Ekaterina Landgren | Ekaterina Landgren, Ryan E. Langendorf and Matthew G. Burgess | Optimal ambition in business, politics and life | 11 pages, 6 figures | null | null | null | physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In business, politics and life, folk wisdom encourages people to aim for above-average results, but to not let the perfect be the enemy of the good. Here, we mathematically formalize and extend this folk wisdom. We model a time-limited search for strategies having uncertain rewards. At each time step, the searcher either is satisfied with their current reward or continues searching. We prove that the optimal satisfaction threshold is both finite and strictly larger than the mean of available rewards -- matching the folk wisdom. We show that being too ambitious has a higher expected cost than being too cautious. We show that the optimal satisfaction threshold increases if the search time is longer, or if the reward distribution is rugged (i.e., has low autocorrelation) or left skewed. The skewness result reveals counterintuitive contrasts between optimal ambition and optimal risk taking. We show that using upward social comparison to assess the reward landscape substantially harms expected performance. We show how these insights can be applied in real-world settings, using examples from entrepreneurship, economic policy, political campaigns, online dating and college admissions. | [
{
"version": "v1",
"created": "Fri, 14 Feb 2025 19:00:04 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Feb 2025 23:34:14 GMT"
},
{
"version": "v3",
"created": "Mon, 26 May 2025 23:40:03 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Landgren",
"Ekaterina",
""
],
[
"Langendorf",
"Ryan E.",
""
],
[
"Burgess",
"Matthew G.",
""
]
] |
2502.10700 | Santanu Maiti K. | Debjani Das Gupta, Santanu K. Maiti | Antiferromagnetic diamond network as an efficient spin filter: Proposition of a spin-specific semi-conducting behavior | 11 pages, 14 figures (Accepted for publication in Annals of Physics) | Annals of Physics 480, 170077 (2025) | 10.1016/j.aop.2025.170077 | null | cond-mat.mes-hall physics.app-ph | http://creativecommons.org/licenses/by/4.0/ | We propose, for the first time, that an array of diamond plaquettes, each possessing vanishing net magnetization, can achieve complete spin polarization over a broad bias window. Furthermore, this system can be utilized to realize spin-specific semiconducting behavior. We describe the antiferromagnetic diamond network within a tight-binding framework, where spin-dependent scattering arises due to the interaction between itinerant electrons and local magnetic moments at different lattice sites. The mechanism underlying spin filtration relies on the specific arrangement of magnetic moments within individual plaquettes. We systematically investigate the spin polarization phenomenon under various input conditions, examining its dependence on network size, system temperature, and the magnetic flux threading each plaquette. Due to the network's geometry, we identify a sharply localized, highly degenerate energy level coexisting with conducting states. By tuning physical parameters, a small energy gap can be established between these degenerate localized states and the conducting energy band, enabling spin-specific $p$-type and $n$-type semiconducting behavior. Our findings offer a novel approach for designing future spintronic devices based on similar antiferromagnetic networks. | [
{
"version": "v1",
"created": "Sat, 15 Feb 2025 07:12:29 GMT"
},
{
"version": "v2",
"created": "Tue, 06 May 2025 11:53:21 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Gupta",
"Debjani Das",
""
],
[
"Maiti",
"Santanu K.",
""
]
] |
2502.10764 | Seokbin Yoon | Hong-ah Chai, Seokbin Yoon, Keumjin Lee | Learning to Explain Air Traffic Situation | 5 pages, 3 figures, minor revisions to address reviewer feedback for final submission to the First US-Europe Air Transportation Research and Development (ATRD) Symposium | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding how air traffic controllers construct a mental 'picture' of complex air traffic situations is crucial but remains a challenge due to the inherently intricate, high-dimensional interactions between aircraft, pilots, and controllers. Previous work on modeling the strategies of air traffic controllers and their mental image of traffic situations often centers on specific air traffic control tasks or pairwise interactions between aircraft, neglecting to capture the comprehensive dynamics of an air traffic situation. To address this issue, we propose a machine learning-based framework for explaining air traffic situations. Specifically, we employ a Transformer-based multi-agent trajectory model that encapsulates both the spatio-temporal movement of aircraft and social interaction between them. By deriving attention scores from the model, we can quantify the influence of individual aircraft on overall traffic dynamics. This provides explainable insights into how air traffic controllers perceive and understand the traffic situation. Trained on real-world air traffic surveillance data collected from the terminal airspace around Incheon International Airport in South Korea, our framework effectively explicates air traffic situations. This could potentially support and enhance the decision-making and situational awareness of air traffic controllers. | [
{
"version": "v1",
"created": "Sat, 15 Feb 2025 11:03:47 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 06:07:27 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chai",
"Hong-ah",
""
],
[
"Yoon",
"Seokbin",
""
],
[
"Lee",
"Keumjin",
""
]
] |
2502.11051 | Jiahao Huo | Jiahao Huo, Yibo Yan, Xu Zheng, Yuanhuiyi Lyu, Xin Zou, Zhihua Wei, Xuming Hu | MMUnlearner: Reformulating Multimodal Machine Unlearning in the Era of Multimodal Large Language Models | Accepted as ACL 2025 Findings | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent progress in Machine Unlearning (MU) has introduced solutions for the selective removal of private or sensitive information encoded within deep neural networks. Nonetheless, MU for Multimodal Large Language Models (MLLMs) remains in its nascent phase. Therefore, we propose to reformulate the task of multimodal MU in the era of MLLMs, which aims to erase only the visual patterns associated with a given entity while preserving the corresponding textual knowledge encoded within the original parameters of the language model backbone. Furthermore, we develop a novel geometry-constrained gradient ascent method MMUnlearner. It updates the weights of MLLMs with a weight saliency map jointly restricted by the remaining concepts and textual knowledge during unlearning, thereby preserving parameters essential for non-target knowledge. Extensive experiments demonstrate that MMUnlearner surpasses baselines that finetuning MLLMs with VQA data directly through Gradient Ascent (GA) or Negative Preference Optimization (NPO), across all evaluation dimensions. Our code can be found in [this URL](https://github.com/Z1zs/MMUnlearner). | [
{
"version": "v1",
"created": "Sun, 16 Feb 2025 09:23:50 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Feb 2025 13:32:25 GMT"
},
{
"version": "v3",
"created": "Tue, 20 May 2025 15:47:22 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 14:55:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Huo",
"Jiahao",
""
],
[
"Yan",
"Yibo",
""
],
[
"Zheng",
"Xu",
""
],
[
"Lyu",
"Yuanhuiyi",
""
],
[
"Zou",
"Xin",
""
],
[
"Wei",
"Zhihua",
""
],
[
"Hu",
"Xuming",
""
]
] |
2502.11198 | Bidyarthi Paul | Bidyarthi Paul, Faika Fairuj Preotee, Shuvashis Sarker, Shamim Rahim Refat, Shifat Islam, Tashreef Muhammad, Mohammad Ashraful Hoque, Shahriar Manzoor | ANCHOLIK-NER: A Benchmark Dataset for Bangla Regional Named Entity Recognition | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Named Entity Recognition (NER) in regional dialects is a critical yet underexplored area in Natural Language Processing (NLP), especially for low-resource languages like Bangla. While NER systems for Standard Bangla have made progress, no existing resources or models specifically address the challenge of regional dialects such as Barishal, Chittagong, Mymensingh, Noakhali, and Sylhet, which exhibit unique linguistic features that existing models fail to handle effectively. To fill this gap, we introduce ANCHOLIK-NER, the first benchmark dataset for NER in Bangla regional dialects, comprising 17,405 sentences distributed across five regions. The dataset was sourced from publicly available resources and supplemented with manual translations, ensuring alignment of named entities across dialects. We evaluate three transformer-based models - Bangla BERT, Bangla BERT Base, and BERT Base Multilingual Cased - on this dataset. Our findings demonstrate that BERT Base Multilingual Cased performs best in recognizing named entities across regions, with significant performance observed in Mymensingh with an F1-score of 82.611%. Despite strong overall performance, challenges remain in region like Chittagong, where the models show lower precision and recall. Since no previous NER systems for Bangla regional dialects exist, our work represents a foundational step in addressing this gap. Future work will focus on improving model performance in underperforming regions and expanding the dataset to include more dialects, enhancing the development of dialect-aware NER systems. | [
{
"version": "v1",
"created": "Sun, 16 Feb 2025 16:59:10 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Mar 2025 14:13:50 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 17:14:48 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Paul",
"Bidyarthi",
""
],
[
"Preotee",
"Faika Fairuj",
""
],
[
"Sarker",
"Shuvashis",
""
],
[
"Refat",
"Shamim Rahim",
""
],
[
"Islam",
"Shifat",
""
],
[
"Muhammad",
"Tashreef",
""
],
[
"Hoque",
"Mohammad Ashraful",
""
],
[
"Manzoor",
"Shahriar",
""
]
] |
2502.11216 | Madhurjya P Bora Dr | Hitendra Sarkar and Madhurjya P. Bora | Exploring nonlinearities in a positive ion-negative ion (PINI) plasma: can other processes mimic debris-induced effects? | 17 pages, 6 figures | null | null | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, an analysis of nonlinear waves and structures induced by an external charged debris in a positive ion-negative ion (PINI) plasma is presented. The results obtained are compared with findings from available experiments involving PINI plasma. The process of formation of different nonlinear structures is examined theoretically through a forced Korteweg-de Vries (fKdV) equation, which is also verified with a multi-fluid flux-corrected transport simulation code mFCT. Various processes which are responsible for different nonlinear waves and structures excited by differently charged external debris are pointed out. This work also points out the similarities in different nonlinear structures {excited} by an external charged debris and the underlying processes (this work) and those {observed experimentally through processes that do not involve any external debris. | [
{
"version": "v1",
"created": "Sun, 16 Feb 2025 17:37:42 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Apr 2025 08:16:08 GMT"
},
{
"version": "v3",
"created": "Thu, 17 Apr 2025 02:31:44 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 03:51:19 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Sarkar",
"Hitendra",
""
],
[
"Bora",
"Madhurjya P.",
""
]
] |
2502.12051 | Ayan Sengupta | Ayan Sengupta, Yash Goel and Tanmoy Chakraborty | How to Upscale Neural Networks with Scaling Law? A Survey and Practical Guidelines | 21 pages, 11 tables, 4 figures | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Neural scaling laws have revolutionized the design and optimization of large-scale AI models by revealing predictable relationships between model size, dataset volume, and computational resources. Early research established power-law relationships in model performance, leading to compute-optimal scaling strategies. However, recent studies highlighted their limitations across architectures, modalities, and deployment contexts. Sparse models, mixture-of-experts, retrieval-augmented learning, and multimodal models often deviate from traditional scaling patterns. Moreover, scaling behaviors vary across domains such as vision, reinforcement learning, and fine-tuning, underscoring the need for more nuanced approaches. In this survey, we synthesize insights from over 50 studies, examining the theoretical foundations, empirical findings, and practical implications of scaling laws. We also explore key challenges, including data efficiency, inference scaling, and architecture-specific constraints, advocating for adaptive scaling strategies tailored to real-world applications. We suggest that while scaling laws provide a useful guide, they do not always generalize across all architectures and training strategies. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 17:20:41 GMT"
},
{
"version": "v2",
"created": "Sat, 24 May 2025 03:19:41 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 03:20:08 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Sengupta",
"Ayan",
""
],
[
"Goel",
"Yash",
""
],
[
"Chakraborty",
"Tanmoy",
""
]
] |
2502.12110 | Wujiang Xu | Wujiang Xu and Kai Mei and Hang Gao and Juntao Tan and Zujie Liang and Yongfeng Zhang | A-MEM: Agentic Memory for LLM Agents | null | null | null | null | cs.CL cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While large language model (LLM) agents can effectively use external tools for complex real-world tasks, they require memory systems to leverage historical experiences. Current memory systems enable basic storage and retrieval but lack sophisticated memory organization, despite recent attempts to incorporate graph databases. Moreover, these systems' fixed operations and structures limit their adaptability across diverse tasks. To address this limitation, this paper proposes a novel agentic memory system for LLM agents that can dynamically organize memories in an agentic way. Following the basic principles of the Zettelkasten method, we designed our memory system to create interconnected knowledge networks through dynamic indexing and linking. When a new memory is added, we generate a comprehensive note containing multiple structured attributes, including contextual descriptions, keywords, and tags. The system then analyzes historical memories to identify relevant connections, establishing links where meaningful similarities exist. Additionally, this process enables memory evolution - as new memories are integrated, they can trigger updates to the contextual representations and attributes of existing historical memories, allowing the memory network to continuously refine its understanding. Our approach combines the structured organization principles of Zettelkasten with the flexibility of agent-driven decision making, allowing for more adaptive and context-aware memory management. Empirical experiments on six foundation models show superior improvement against existing SOTA baselines. The source code for evaluating performance is available at https://github.com/WujiangXu/AgenticMemory, while the source code of agentic memory system is available at https://github.com/agiresearch/A-mem. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 18:36:14 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Mar 2025 04:14:02 GMT"
},
{
"version": "v3",
"created": "Tue, 04 Mar 2025 15:09:10 GMT"
},
{
"version": "v4",
"created": "Mon, 14 Apr 2025 15:21:49 GMT"
},
{
"version": "v5",
"created": "Fri, 18 Apr 2025 17:26:57 GMT"
},
{
"version": "v6",
"created": "Sun, 11 May 2025 18:10:25 GMT"
},
{
"version": "v7",
"created": "Wed, 21 May 2025 05:16:32 GMT"
},
{
"version": "v8",
"created": "Tue, 27 May 2025 02:44:13 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Xu",
"Wujiang",
""
],
[
"Mei",
"Kai",
""
],
[
"Gao",
"Hang",
""
],
[
"Tan",
"Juntao",
""
],
[
"Liang",
"Zujie",
""
],
[
"Zhang",
"Yongfeng",
""
]
] |
2502.12134 | Yige Xu | Yige Xu, Xu Guo, Zhiwei Zeng, Chunyan Miao | SoftCoT: Soft Chain-of-Thought for Efficient Reasoning with LLMs | Camera-ready for ACL 2025 (main conference) | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Chain-of-Thought (CoT) reasoning enables Large Language Models (LLMs) to solve complex reasoning tasks by generating intermediate reasoning steps. However, most existing approaches focus on hard token decoding, which constrains reasoning within the discrete vocabulary space and may not always be optimal. While recent efforts explore continuous-space reasoning, they often require full-model fine-tuning and suffer from catastrophic forgetting, limiting their applicability to state-of-the-art LLMs that already perform well in zero-shot settings with a proper instruction. To address this challenge, we propose a novel approach for continuous-space reasoning that does not require modifying the LLM. Specifically, we employ a lightweight fixed assistant model to speculatively generate instance-specific soft thought tokens as the initial chain of thoughts, which are then mapped into the LLM's representation space via a trainable projection module. Experimental results on five reasoning benchmarks demonstrate that our method enhances LLM reasoning performance through supervised, parameter-efficient fine-tuning. Source code is available at https://github.com/xuyige/SoftCoT. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 18:52:29 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 14:54:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Xu",
"Yige",
""
],
[
"Guo",
"Xu",
""
],
[
"Zeng",
"Zhiwei",
""
],
[
"Miao",
"Chunyan",
""
]
] |
2502.12187 | Atsushi Suzuki | Atsushi Suzuki, Yulan He, Feng Tian, Zhongyuan Wang | Hallucinations are inevitable but can be made statistically negligible. The "innate" inevitability of hallucinations cannot explain practical LLM issues | null | null | null | null | cs.CL cs.FL cs.LG math.ST stat.ML stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hallucinations, a phenomenon where a language model (LM) generates nonfactual content, pose a significant challenge to the practical deployment of LMs. While many empirical methods have been proposed to mitigate hallucinations, recent studies established a computability-theoretic result showing that any LM will inevitably generate hallucinations on an infinite set of inputs, regardless of the quality and quantity of training datasets and the choice of the language model architecture and training and inference algorithms. Although the computability-theoretic result may seem pessimistic, its significance in practical viewpoints has remained unclear. This paper claims that those "innate" inevitability results from computability theory and diagonal argument, in principle, cannot explain practical issues of LLMs. We demonstrate this claim by presenting a positive theoretical result from a probabilistic perspective. Specifically, we prove that hallucinations can be made statistically negligible, provided that the quality and quantity of the training data are sufficient. Interestingly, our positive result coexists with the computability-theoretic result, implying that while hallucinations on an infinite set of inputs cannot be entirely eliminated, their probability can always be reduced by improving algorithms and training data. By evaluating the two seemingly contradictory results through the lens of information theory, we argue that our probability-theoretic positive result better reflects practical considerations than the computability-theoretic negative result. | [
{
"version": "v1",
"created": "Sat, 15 Feb 2025 07:28:40 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 12:53:34 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Suzuki",
"Atsushi",
""
],
[
"He",
"Yulan",
""
],
[
"Tian",
"Feng",
""
],
[
"Wang",
"Zhongyuan",
""
]
] |
2502.12239 | Nikolay Britavskiy | N. Britavskiy, L. Mahy, D. J. Lennon, L. R. Patrick, H. Sana, J. I. Villase\~nor, T. Shenar, J. Bodensteiner, M. Bernini-Peron, S. R. Berlanas, D. M. Bowman, P. A. Crowther, S. E. de Mink, C. J. Evans, Y. G\"otberg, G. Holgado, C. Johnston, Z. Keszthelyi, J. Klencki, N. Langer, I. Mandel, A. Menon, M. Moe, L. M. Oskinova, D. Pauli, M. Pawlak, V. Ramachandran, M. Renzo, A. A. C. Sander, F. R. N. Schneider, A. Schootemeijer, K. Sen, S. Sim\'on-D\'iaz, J. Th. van Loon, J. S. Vink | Binarity at LOw Metallicity (BLOeM): Multiplicity of early B-type supergiants in the Small Magellanic Cloud | 18 pages, 14 figures, accepted for publication in Astronomy & Astrophysics | A&A 698, A40 (2025) | 10.1051/0004-6361/202452963 | null | astro-ph.SR astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The blue supergiant (BSG) domain contains a large variety of stars whose past and future evolutionary paths are still highly uncertain. Since binary interaction plays a crucial role in the fate of massive stars, investigating the multiplicity among BSGs helps shed light on the fate of such objects. We aim to estimate the binary fraction of a large sample of BSGs in the Small Magellanic Cloud within the Binarity at LOw Metallicity (BLOeM) survey. In total, we selected 262 targets with spectral types B0-B3 and luminosity classes I-II. This work is based on spectroscopic data collected by the GIRAFFE instrument, mounted on the Very Large Telescope, which gathered nine epochs over three months. Our spectroscopic analysis for each target includes the individual and peak-to-peak radial velocity measurements, an investigation of the line profile variability, and a periodogram analysis to search for possible short- and long-period binaries. By applying a 20 km s$^{-1}$ threshold on the peak-to-peak radial velocities above which we would consider the star to be binary, the resulting observed spectroscopic binary fraction for our BSG sample is 23 $\pm$ 3$\%$. In addition, we derived reliable orbital periods for 41 spectroscopic binaries and potential binary candidates, among which there are 17 eclipsing binaries, including 20 SB1 and SB2 systems with periods of less than 10 days. We reported a significant drop in the binary fraction of BSGs with spectral types later than B2 and effective temperatures less than 18 kK, which could indicate the end of the main sequence phase in this temperature regime. We found no metallicity dependence in the binary fraction of BSGs, compared to existing spectroscopic surveys of the Galaxy and Large Magellanic Cloud. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 19:00:01 GMT"
},
{
"version": "v2",
"created": "Mon, 28 Apr 2025 15:32:53 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Britavskiy",
"N.",
""
],
[
"Mahy",
"L.",
""
],
[
"Lennon",
"D. J.",
""
],
[
"Patrick",
"L. R.",
""
],
[
"Sana",
"H.",
""
],
[
"Villaseñor",
"J. I.",
""
],
[
"Shenar",
"T.",
""
],
[
"Bodensteiner",
"J.",
""
],
[
"Bernini-Peron",
"M.",
""
],
[
"Berlanas",
"S. R.",
""
],
[
"Bowman",
"D. M.",
""
],
[
"Crowther",
"P. A.",
""
],
[
"de Mink",
"S. E.",
""
],
[
"Evans",
"C. J.",
""
],
[
"Götberg",
"Y.",
""
],
[
"Holgado",
"G.",
""
],
[
"Johnston",
"C.",
""
],
[
"Keszthelyi",
"Z.",
""
],
[
"Klencki",
"J.",
""
],
[
"Langer",
"N.",
""
],
[
"Mandel",
"I.",
""
],
[
"Menon",
"A.",
""
],
[
"Moe",
"M.",
""
],
[
"Oskinova",
"L. M.",
""
],
[
"Pauli",
"D.",
""
],
[
"Pawlak",
"M.",
""
],
[
"Ramachandran",
"V.",
""
],
[
"Renzo",
"M.",
""
],
[
"Sander",
"A. A. C.",
""
],
[
"Schneider",
"F. R. N.",
""
],
[
"Schootemeijer",
"A.",
""
],
[
"Sen",
"K.",
""
],
[
"Simón-Díaz",
"S.",
""
],
[
"van Loon",
"J. Th.",
""
],
[
"Vink",
"J. S.",
""
]
] |
2502.12467 | Jeremy Watson | Jeremy D. Watson | Hybrid Data-enabled Predictive Control: Incorporating model knowledge into the DeePC | 8 pages, 3 figures. Accompanying code repository: https://github.com/jerrydonaldwatson/HDeePC | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Predictive control can either be data-based (e.g. data-enabled predictive control, or DeePC) or model-based (model predictive control). In this paper we aim to bridge the gap between the two by investigating the case where only a partial model is available, i.e. incorporating model knowledge into DeePC. This has potential advantages over a purely data-based approach in terms of noise and computational expense in some cases, as well as applications to certain linear time-varying and nonlinear systems. We formulate an approach to take advantage of partial model knowledge which we call hybrid data-enabled predictive control (HDeePC) and prove feasible set equivalence and equivalent closed-loop behavior in the noiseless, LTI case. Finally, two examples illustrate the potential of HDeePC. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 02:54:42 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Apr 2025 00:55:05 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 09:08:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Watson",
"Jeremy D.",
""
]
] |
2502.12635 | Yunhao Gou | Yunhao Gou, Hansi Yang, Zhili Liu, Kai Chen, Yihan Zeng, Lanqing Hong, Zhenguo Li, Qun Liu, Bo Han, James T. Kwok, Yu Zhang | Corrupted but Not Broken: Understanding and Mitigating the Negative Impacts of Corrupted Data in Visual Instruction Tuning | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Visual Instruction Tuning (VIT) aims to enhance Multimodal Large Language Models (MLLMs), yet its effectiveness is often compromised by corrupted datasets with issues such as hallucinated content, incorrect responses, and poor OCR quality. Previous approaches to address these challenges have focused on refining datasets through high-quality data collection or rule-based filtering that can be costly or limited in scope. In this paper, we conduct a systematic investigation into the impact of corrupted data on MLLMs and discover that, although corrupted data degrade model performance, such adverse effects are largely reversible, and MLLMs are {\bf corrupted but not broken}. Specifically, we find that disabling a small subset of parameters can almost fully restore performance. Moreover, corrupted MLLMs inherently possess the capability to differentiate between clean and corrupted samples, facilitating dataset cleaning without external intervention. Building on these insights, we introduce a corruption-robust training paradigm that significantly surpasses existing strategies for mitigating the effects of corrupted data. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 08:28:29 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 03:24:41 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 08:32:23 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Gou",
"Yunhao",
""
],
[
"Yang",
"Hansi",
""
],
[
"Liu",
"Zhili",
""
],
[
"Chen",
"Kai",
""
],
[
"Zeng",
"Yihan",
""
],
[
"Hong",
"Lanqing",
""
],
[
"Li",
"Zhenguo",
""
],
[
"Liu",
"Qun",
""
],
[
"Han",
"Bo",
""
],
[
"Kwok",
"James T.",
""
],
[
"Zhang",
"Yu",
""
]
] |
2502.12696 | Ignacio Esteban Lopez Delgado | Ignacio E. L\'opez-Delgado, V\'ictor Navarro-L\'opez, Francisco Grandas-P\'erez, Juan I. Godino-Llorente, Jes\'us Grajal | Radar Network for Gait Monitoring: Technology and Validation | null | null | null | null | eess.SP cs.SY eess.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In recent years, radar-based devices have emerged as an alternative approach for gait monitoring. However, the radar configuration and the algorithms used to extract the gait parameters often differ between contributions, lacking a systematic evaluation of the most appropriate setup. Additionally, radar-based studies often exclude motorically impaired subjects, leaving it unclear whether the existing algorithms are applicable to such populations. In this paper, a radar network is developed and validated by monitoring the gait of five healthy individuals and three patients with Parkinson's disease. Six configurations and four algorithms were compared using Vicon as ground-truth to determine the most appropriate solution for gait monitoring. The best results were obtained using only three nodes: two oriented towards the feet and one towards the torso. The most accurate stride velocity and distance in the state of the art were obtained with this configuration. Moreover, we show that analyzing the feet velocity increases the reliability of the temporal parameters, especially with aged or motorically impaired subjects. The contribution is significant for the implementation of radar networks in clinical and domestic environments, as it addresses critical aspects concerning the radar network configuration and algorithms. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 09:54:13 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 10:10:36 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"López-Delgado",
"Ignacio E.",
""
],
[
"Navarro-López",
"Víctor",
""
],
[
"Grandas-Pérez",
"Francisco",
""
],
[
"Godino-Llorente",
"Juan I.",
""
],
[
"Grajal",
"Jesús",
""
]
] |
2502.12706 | Jing Xu | Jing Xu, Jiazheng Li, Jingzhao Zhang | Scalable Model Merging with Progressive Layer-wise Distillation | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Model merging offers an effective way to integrate the capabilities of multiple fine-tuned models. However, the performance degradation of the merged model remains a challenge, particularly when none or few data are available. This paper first highlights the necessity of domain-specific data for model merging by proving that data-agnostic algorithms can have arbitrarily bad worst-case performance. Building on this theoretical insight, we explore the relationship between model merging and distillation, introducing a novel few-shot merging algorithm, ProDistill (Progressive Layer-wise Distillation). Unlike common belief that layer wise training hurts performance, we show that layer-wise teacher-student distillation not only enhances the scalability but also improves model merging performance. We conduct extensive experiments to show that compared to existing few-shot merging methods, ProDistill achieves state-of-the-art performance, with up to 6.14% and 6.61% improvements in vision and NLU tasks. Furthermore, we extend the experiments to models with over 10B parameters, showcasing the exceptional scalability of ProDistill. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 10:15:18 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 05:56:07 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Xu",
"Jing",
""
],
[
"Li",
"Jiazheng",
""
],
[
"Zhang",
"Jingzhao",
""
]
] |
2502.12777 | Niladri Sett | Bhargavi Kalyani I, A Rama Prasad Mathi, Niladri Sett | Evaluating link prediction: New perspectives and recommendations | null | null | null | null | cs.SI cs.AI | http://creativecommons.org/licenses/by/4.0/ | Link prediction (LP) is an important problem in network science and machine learning research. The state-of-the-art LP methods are usually evaluated in a uniform setup, ignoring several factors associated with the data and application specific needs. We identify a number of such factors, such as, network-type, problem-type, geodesic distance between the end nodes and its distribution over the classes, nature and applicability of LP methods, class imbalance and its impact on early retrieval, evaluation metric, etc., and present an experimental setup which allows us to evaluate LP methods in a rigorous and controlled manner. We perform extensive experiments with a variety of LP methods over real network datasets in this controlled setup, and gather valuable insights on the interactions of these factors with the performance of LP through an array of carefully designed hypotheses. Following the insights, we provide recommendations to be followed as best practice for evaluating LP methods. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 11:36:59 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 06:28:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kalyani",
"Bhargavi",
"I"
],
[
"Mathi",
"A Rama Prasad",
""
],
[
"Sett",
"Niladri",
""
]
] |
2502.12845 | Nian Ran | Nian Ran, Yue Wang, Richard Allmendinger | MOLLM: Multi-Objective Large Language Model for Molecular Design -- Optimizing with Experts | 9 pages, under review | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Molecular design plays a critical role in advancing fields such as drug discovery, materials science, and chemical engineering. This work introduces the Multi-Objective Large Language Model for Molecular Design (MOLLM), a novel framework that combines domain-specific knowledge with the adaptability of large language models to optimize molecular properties across multiple objectives. Leveraging in-context learning and multi-objective optimization, MOLLM achieves superior performance and innovation, consistently surpassing state-of-the-art (SOTA) methods. We significantly improve the efficiency of our framework, making it 14 times faster and substantially more cost-effective without compromising performance compared to the latest similar work. Our results demonstrate that MOLLM consistently outperforms SOTA models across experiments and excels on the PMO benchmark. In addition, we provide extensive ablation studies and analysis to evaluate the effectiveness of each component and the quality of the output molecules. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 13:25:00 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 05:36:24 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ran",
"Nian",
""
],
[
"Wang",
"Yue",
""
],
[
"Allmendinger",
"Richard",
""
]
] |
2502.12953 | Radu Tudor Ionescu | Andrei Jarca, Florinel Alin Croitoru, Radu Tudor Ionescu | Task-Informed Anti-Curriculum by Masking Improves Downstream Performance on Text | Accepted at ACL 2025 | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Masked language modeling has become a widely adopted unsupervised technique to pre-train large language models (LLMs). However, the process of selecting tokens for masking is random, and the percentage of masked tokens is typically fixed for the entire training process. In this paper, we propose to adjust the masking ratio and to decide which tokens to mask based on a novel task-informed anti-curriculum learning scheme. First, we harness task-specific knowledge about useful and harmful tokens in order to determine which tokens to mask. Second, we propose a cyclic decaying masking ratio, which corresponds to an anti-curriculum schedule (from hard to easy). We exemplify our novel task-informed anti-curriculum by masking (TIACBM) approach across three diverse downstream tasks: sentiment analysis, text classification by topic, and authorship attribution. Our findings suggest that TIACBM enhances the ability of the model to focus on key task-relevant features, contributing to statistically significant performance gains across tasks. We release our code at https://github.com/JarcaAndrei/TIACBM. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 15:36:16 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 22:37:20 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Jarca",
"Andrei",
""
],
[
"Croitoru",
"Florinel Alin",
""
],
[
"Ionescu",
"Radu Tudor",
""
]
] |
2502.13010 | Mohammad R. Rezaei | Mohammad Reza Rezaei, Reza Saadati Fard, Rahul G. Krishnan, Milad Lankarany | Agentic Medical Knowledge Graphs Enhance Medical Question Answering: Bridging the Gap Between LLMs and Evolving Medical Knowledge | null | null | null | null | cs.CL cs.MA | http://creativecommons.org/licenses/by-sa/4.0/ | Large Language Models (LLMs) have significantly advanced medical question-answering by leveraging extensive clinical data and medical literature. However, the rapid evolution of medical knowledge and the labor-intensive process of manually updating domain-specific resources pose challenges to the reliability of these systems. To address this, we introduce Agentic Medical Graph-RAG (AMG-RAG), a comprehensive framework that automates the construction and continuous updating of medical knowledge graphs, integrates reasoning, and retrieves current external evidence, such as PubMed and WikiSearch. By dynamically linking new findings and complex medical concepts, AMG-RAG not only improves accuracy but also enhances interpretability in medical queries.
Evaluations on the MEDQA and MEDMCQA benchmarks demonstrate the effectiveness of AMG-RAG, achieving an F1 score of 74.1 percent on MEDQA and an accuracy of 66.34 percent on MEDMCQA, outperforming both comparable models and those 10 to 100 times larger. Notably, these improvements are achieved without increasing computational overhead, highlighting the critical role of automated knowledge graph generation and external evidence retrieval in delivering up-to-date, trustworthy medical insights. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 16:29:45 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:05:15 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Rezaei",
"Mohammad Reza",
""
],
[
"Fard",
"Reza Saadati",
""
],
[
"Krishnan",
"Rahul G.",
""
],
[
"Lankarany",
"Milad",
""
]
] |
2502.13398 | Vishal Dey | Vishal Dey, Xiao Hu, Xia Ning | GeLLMO: Generalizing Large Language Models for Multi-property Molecule Optimization | Accepted to ACL Main 2025. Vishal Dey and Xiao Hu contributed equally to this paper | null | null | null | cs.LG cs.AI cs.CL physics.chem-ph q-bio.QM | http://creativecommons.org/licenses/by/4.0/ | Despite recent advancements, most computational methods for molecule optimization are constrained to single- or double-property optimization tasks and suffer from poor scalability and generalizability to novel optimization tasks. Meanwhile, Large Language Models (LLMs) demonstrate remarkable out-of-domain generalizability to novel tasks. To demonstrate LLMs' potential for molecule optimization, we introduce MuMOInstruct, the first high-quality instruction-tuning dataset specifically focused on complex multi-property molecule optimization tasks. Leveraging MuMOInstruct, we develop GeLLMOs, a series of instruction-tuned LLMs for molecule optimization. Extensive evaluations across 5 in-domain and 5 out-of-domain tasks demonstrate that GeLLMOs consistently outperform state-of-the-art baselines. GeLLMOs also exhibit outstanding zero-shot generalization to unseen tasks, significantly outperforming powerful closed-source LLMs. Such strong generalizability demonstrates the tremendous potential of GeLLMOs as foundational models for molecule optimization, thereby tackling novel optimization tasks without resource-intensive retraining. MuMOInstruct, models, and code are accessible through https://github.com/ninglab/GeLLMO. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 03:14:11 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:37:58 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Dey",
"Vishal",
""
],
[
"Hu",
"Xiao",
""
],
[
"Ning",
"Xia",
""
]
] |
2502.13600 | Shunsuke Kurima Mr. | Pierluigi Colli, Shunsuke Kurima | Analysis of a nonisothermal and conserved phase field system with inertial term | Key words and phrases: conserved phase field system; inertial term; non-isothermal process; nonlinear partial differential equations; initial-boundary value problem; existence of solutions | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper deals with a conserved phase field system that couples the energy balance equation with a Cahn--Hilliard type system including temperature and the inertial term for the order parameter. In the case without inertial term, the system under study was introduced by Caginalp. The inertial term is motivated by the occurrence of rapid phase transformation processes in nonequilibrium dynamics. A double-well potential is well chosen and the related nonlinearity governing the evolution is assumed to satisfy a suitable growth condition. The viscous variant of the Cahn--Hilliard system is also considered along with the inertial term. The existence of a global solution is proved via the analysis of some approximate problems with Yosida regularizations, and the use of the Cauchy--Lipschitz--Picard theorem in an abstract setting. Moreover, we study the convergence of the system, with or without the viscous term, as the inertial coefficient tends to zero. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 10:24:26 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 08:23:09 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Colli",
"Pierluigi",
""
],
[
"Kurima",
"Shunsuke",
""
]
] |
2502.14132 | Nadav Borenstein | Nadav Borenstein, Greta Warren, Desmond Elliott, Isabelle Augenstein | Can Community Notes Replace Professional Fact-Checkers? | Accepted to the main proceedings of ACL 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Two commonly employed strategies to combat the rise of misinformation on social media are (i) fact-checking by professional organisations and (ii) community moderation by platform users. Policy changes by Twitter/X and, more recently, Meta, signal a shift away from partnerships with fact-checking organisations and towards an increased reliance on crowdsourced community notes. However, the extent and nature of dependencies between fact-checking and helpful community notes remain unclear. To address these questions, we use language models to annotate a large corpus of Twitter/X community notes with attributes such as topic, cited sources, and whether they refute claims tied to broader misinformation narratives. Our analysis reveals that community notes cite fact-checking sources up to five times more than previously reported. Fact-checking is especially crucial for notes on posts linked to broader narratives, which are twice as likely to reference fact-checking sources compared to other sources. Our results show that successful community moderation relies on professional fact-checking and highlight how citizen and professional fact-checking are deeply intertwined. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 22:26:39 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 10:54:58 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Borenstein",
"Nadav",
""
],
[
"Warren",
"Greta",
""
],
[
"Elliott",
"Desmond",
""
],
[
"Augenstein",
"Isabelle",
""
]
] |
2502.14213 | Yanchen Yin | Yanchen Yin and Yongli Wang | Communication-Efficient Solving of Linear Systems via Asynchronous Randomized Block Projections in ROS2 Networks | 20 pages, 2 figures | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes an event-triggered asynchronous distributed randomized block Kaczmarz projection (ER-AD-RBKP) algorithm for efficiently solving large-scale linear systems in resource-constrained and communication-unstable environments. The algorithm enables each agent to update its local state estimate independently and engage in communication only when specific triggering conditions are satisfied, thereby significantly reducing communication overhead. At each iteration, agents project onto randomized subsets of local data blocks to lower computational cost and enhance scalability.
From a theoretical standpoint, we establish exponential convergence conditions for the proposed algorithm. By defining events that ensure strong connectivity in the communication graph, we derive sufficient conditions for global convergence under a probabilistic framework. Our analysis guarantees that the algorithm is guaranteed to converge in expectation under mild probabilistic assumptions, provided that persistent agent isolation is avoided. For inconsistent systems, auxiliary variables are incorporated to transform the problem into an equivalent consistent formulation, and theoretical error bounds are derived.
For practical evaluation, we implement the ER-AD-RBKP algorithm in an asynchronous communication environment built on ROS2, a distributed middleware framework for real-time robotic systems. We evaluate the algorithm under various settings, including varying numbers of agents, neighborhood sizes, communication intervals, and failure scenarios such as communication disruptions and processing faults. Experimental results demonstrate robust performance in terms of computational efficiency, communication cost, and system resilience, highlighting its strong potential for practical applicability in real-world distributed systems. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 02:47:12 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:30:00 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Yin",
"Yanchen",
""
],
[
"Wang",
"Yongli",
""
]
] |
2502.14379 | Hao Qin | Hao Qin, Kwang-Sung Jun, Chicheng Zhang | Achieving adaptivity and optimality for multi-armed bandits using Exponential-Kullback Leibler Maillard Sampling | 10 pages of the main body, 2 figures, 37 pages in total | null | null | null | cs.LG cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problem of $K$-armed bandits with reward distributions belonging to a one-parameter exponential distribution family. In the literature, several criteria have been proposed to evaluate the performance of such algorithms, including Asymptotic Optimality, Minimax Optimality, Sub-UCB, and variance-adaptive worst-case regret bound. Thompson Sampling-based and Upper Confidence Bound-based algorithms have been employed to achieve some of these criteria. However, none of these algorithms simultaneously satisfy all the aforementioned criteria.
In this paper, we design an algorithm, Exponential Kullback-Leibler Maillard Sampling (abbrev. Exp-KL-MS), that can achieve multiple optimality criteria simultaneously, including Asymptotic Optimality, Minimax Optimality with a $\sqrt{\ln (K)}$ factor, Sub-UCB, and variance-adaptive worst-case regret bound. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 09:12:16 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 22:15:21 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Qin",
"Hao",
""
],
[
"Jun",
"Kwang-Sung",
""
],
[
"Zhang",
"Chicheng",
""
]
] |
2502.14479 | Arno Botha | Arno Botha, Tanja Verster, Roland Breedt | Modelling the term-structure of default risk under IFRS 9 within a multistate regression framework | 32 pages, 8211 words, 11 figures | null | null | null | q-fin.RM q-fin.ST stat.AP | http://creativecommons.org/licenses/by/4.0/ | The lifetime behaviour of loans is notoriously difficult to model, which can compromise a bank's financial reserves against future losses, if modelled poorly. Therefore, we present a data-driven comparative study amongst three techniques in modelling a series of default risk estimates over the lifetime of each loan, i.e., its term-structure. The behaviour of loans can be described using a nonstationary and time-dependent semi-Markov model, though we model its elements using a multistate regression-based approach. As such, the transition probabilities are explicitly modelled as a function of a rich set of input variables, including macroeconomic and loan-level inputs. Our modelling techniques are deliberately chosen in ascending order of complexity: 1) a Markov chain; 2) beta regression; and 3) multinomial logistic regression. Using residential mortgage data, our results show that each successive model outperforms the previous, likely as a result of greater sophistication. This finding required devising a novel suite of simple model diagnostics, which can itself be reused in assessing sampling representativeness and the performance of other modelling techniques. These contributions surely advance the current practice within banking when conducting multistate modelling. Consequently, we believe that the estimation of loss reserves will be more timeous and accurate under IFRS 9. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 11:53:20 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:21:53 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Botha",
"Arno",
""
],
[
"Verster",
"Tanja",
""
],
[
"Breedt",
"Roland",
""
]
] |
2502.14488 | Ragnar Groot Koerkamp | Lorraine A. K. Ayad, Gabriele Fici, Ragnar Groot Koerkamp, Grigorios Loukides, Rob Patro, Giulio Ermanno Pibiri, Solon P. Pissis | U-index: A Universal Indexing Framework for Matching Long Patterns | SEA-2025 version. 18 pages, 6 figures, code available at https://github.com/u-index/u-index-rs | null | null | null | cs.DS | http://creativecommons.org/licenses/by/4.0/ | Text indexing is a fundamental and well-studied problem. Classic solutions either replace the original text with a compressed representation, e.g., the FM-index and its variants, or keep it uncompressed but attach some redundancy - an index - to accelerate matching. The former solutions thus retain excellent compressed space, but are slow in practice. The latter approaches, like the suffix array, instead sacrifice space for speed.
We show that efficient text indexing can be achieved using just a small extra space on top of the original text, provided that the query patterns are sufficiently long. More specifically, we develop a new indexing paradigm in which a sketch of a query pattern is first matched against a sketch of the text. Once candidate matches are retrieved, they are verified using the original text. This paradigm is thus universal in the sense that it allows us to use any solution to index the sketched text, like a suffix array, FM-index, or r-index.
We explore both the theory and the practice of this universal framework. With an extensive experimental analysis, we show that, surprisingly, universal indexes can be constructed much faster than their unsketched counterparts and take a fraction of the space, as a direct consequence of (i) having a lower bound on the length of patterns and (ii) working in sketch space. Furthermore, these data structures have the potential of retaining or even improving query time, because matching against the sketched text is faster and verifying candidates can be theoretically done in constant time per occurrence (or, in practice, by short and cache-friendly scans of the text). Finally, we discuss some important applications of this novel indexing paradigm to computational biology. We hypothesize that such indexes will be particularly effective when the queries are sufficiently long, and so demonstrate applications in long-read mapping. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 12:09:34 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Feb 2025 13:35:43 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 12:05:04 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ayad",
"Lorraine A. K.",
""
],
[
"Fici",
"Gabriele",
""
],
[
"Koerkamp",
"Ragnar Groot",
""
],
[
"Loukides",
"Grigorios",
""
],
[
"Patro",
"Rob",
""
],
[
"Pibiri",
"Giulio Ermanno",
""
],
[
"Pissis",
"Solon P.",
""
]
] |
2502.14613 | Jan Trienes | Jan Trienes, J\"org Schl\"otterer, Junyi Jessy Li, Christin Seifert | Behavioral Analysis of Information Salience in Large Language Models | Accepted at ACL 2025 (Findings) | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) excel at text summarization, a task that requires models to select content based on its importance. However, the exact notion of salience that LLMs have internalized remains unclear. To bridge this gap, we introduce an explainable framework to systematically derive and investigate information salience in LLMs through their summarization behavior. Using length-controlled summarization as a behavioral probe into the content selection process, and tracing the answerability of Questions Under Discussion throughout, we derive a proxy for how models prioritize information. Our experiments on 13 models across four datasets reveal that LLMs have a nuanced, hierarchical notion of salience, generally consistent across model families and sizes. While models show highly consistent behavior and hence salience patterns, this notion of salience cannot be accessed through introspection, and only weakly correlates with human perceptions of information salience. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 14:52:23 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 13:13:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Trienes",
"Jan",
""
],
[
"Schlötterer",
"Jörg",
""
],
[
"Li",
"Junyi Jessy",
""
],
[
"Seifert",
"Christin",
""
]
] |
2502.15084 | Zexin Ye | Zexin Ye | Algorithmic Collusion under Observed Demand Shocks | null | null | null | null | econ.GN q-fin.EC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The growing reliance on AI-driven pricing algorithms has raised antitrust concerns, particularly that third-party use of nonpublic competitor data may facilitate information sharing and undermine market competition. This paper examines a setting in which firms receive demand information from a common third party and incorporate it into their pricing algorithms. Specifically, I analyze the pricing behavior of Q-learning agents when demand shocks are observable. The simulation results show that at high discount factors, agents consistently charge constant supracompetitive prices across demand states, resulting in price rigidity. This pricing pattern is sustained through collusive strategies that effectively punish deviations and feature a discontinuous switch from punishment back to cooperation. At medium discount factors, however, agents lower prices during booms to reduce incentives to deviate and raise them during downturns, resulting in countercyclical pricing consistent with Rotemberg and Saloner (1986). The findings highlight both the pricing adaptability of Q-learning to fluctuating market demand and the robustness of algorithmic collusion. Compared to settings with unobserved demand shocks, demand observability reduces firm profits and enhances consumer surplus, suggesting that access to market information under algorithmic pricing does not necessarily harm consumers. | [
{
"version": "v1",
"created": "Thu, 20 Feb 2025 22:51:48 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 19:53:28 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ye",
"Zexin",
""
]
] |
2502.15254 | Hinako Asaoka | Hinako Asaoka, Kazue Kudo | Quantum autoencoders for image classification | null | null | null | null | quant-ph cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Classical machine learning often struggles with complex, high-dimensional data. Quantum machine learning offers a potential solution, promising more efficient processing. The quantum convolutional neural network (QCNN), a hybrid algorithm, fits current noisy intermediate-scale quantum hardware. However, its training depends largely on classical computation. Future gate-based quantum computers may realize full quantum advantages. In contrast to QCNNs, quantum autoencoders (QAEs) leverage classical optimization solely for parameter tuning. Data compression and reconstruction are handled entirely within quantum circuits, enabling purely quantum-based feature extraction. This study introduces a novel image-classification approach using QAEs, achieving classification without requiring additional qubits compared with conventional QAE implementations. The quantum circuit structure significantly impacts classification accuracy. Unlike hybrid methods such as QCNN, QAE-based classification emphasizes quantum computation. Our experiments demonstrate high accuracy in a four-class classification task, evaluating various quantum-gate configurations to understand the impact of different parameterized quantum circuit structures on classification performance. Specifically, noise-free conditions are considered, and simulations are performed using a statevector simulator to model the quantum system with full amplitude precision. Our results reveal that specific ansatz structures achieve superior accuracy. Moreover, the proposed approach achieves performance comparable to that of conventional machine-learning methods while significantly reducing the number of parameters requiring optimization. These findings indicate that QAEs can serve as efficient classification models with fewer parameters and highlight the potential of utilizing quantum circuits for complete end-to-end learning. | [
{
"version": "v1",
"created": "Fri, 21 Feb 2025 07:13:38 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 14:30:37 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Asaoka",
"Hinako",
""
],
[
"Kudo",
"Kazue",
""
]
] |
2502.15539 | Ragnar Groot Koerkamp | Ragnar Groot Koerkamp | PtrHash: Minimal Perfect Hashing at RAM Throughput | SEA-2025 version. 21 pages; 7 figures; source code at https://github.com/RagnarGrootKoerkamp/ptrhash | null | null | null | cs.DS | http://creativecommons.org/licenses/by/4.0/ | Given a set $K$ of $n$ keys, a minimal perfect hash function (MPHF) is a collision-free bijective map $\mathsf{H_{mphf}}$ from $K$ to $\{0, \dots, n-1\}$. This work presents a (minimal) perfect hash function that first prioritizes query throughput, while also allowing efficient construction for $10^9$ or more elements using 2.4 bits of memory per key.
Both PTHash and PHOBIC first map all $n$ keys to $n/\lambda < n$ buckets. Then, each bucket stores a pilot that controls the final hash value of the keys mapping to it. PtrHash builds on this by using 1) fixed-width (uncompressed) 8-bit pilots, 2) a construction algorithm similar to cuckoo-hashing to find suitable pilot values. Further, it 3) uses the same number of buckets and slots for each part, with 4) a single remap table to map intermediate positions $\geq n$ to $<n$, 5) encoded using per-cacheline Elias-Fano coding. Lastly, 6) PtrHash support streaming queries, where we use prefetching to answer a stream of multiple queries more efficiently than one-by-one processing.
With default parameters, PtrHash takes 2.0 bits per key. On 300 million string keys, PtrHash is as fast or faster to build than other MPHFs, and at least $2.1\times$ faster to query. When streaming multiple queries, this improves to $3.3\times$ speedup over the fastest alternative, while also being significantly faster to construct. When using $10^9$ integer keys instead, query times are as low as 12 ns/key when iterating in a for loop, or even down to 8 ns/key when using the streaming approach, just short of the 7.4 ns inverse throughput of random memory accesses. | [
{
"version": "v1",
"created": "Fri, 21 Feb 2025 15:42:18 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 12:30:08 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Koerkamp",
"Ragnar Groot",
""
]
] |
2502.16231 | Apostolos Christou | Apostolos A. Christou and Nikos Georgakarakos and Matija Cuk and Aldo Dell'Oro and Andrew Marshall-Lee and Alice Humpage | New asteroid clusters and evidence of collisional fragmentation in the L5 Trojan cloud of Mars | 10 pages,9 figures, accepted in A&A with minor corrections | A&A 698, A42 (2025) | 10.1051/0004-6361/202553804 | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Context. Trojan asteroids of Mars date from an early phase of solar system evolution. Based on a sample of <10 asteroids, MT distribution has been previously shown to be both asymmetric and inhomogeneous with population evolution dominated by thermal radiation forces acting over timescales of Gyr. Remarkably, a single asteroid family associated with (5261) Eureka (H~16) at L5 contains most MTs and all members of the stable population fainter than H=18.
Aims. Using the currently available sample of MTs and their orbits, we take a fresh look at this population to re-evaluate these earlier conclusions and to search for additional features diagnostic of the evolutionary history of MTs and the Eureka family in particular.
Methods. We perform harmonic analysis on numerical time series of the osculating elements to compile a new proper element catalogue comprising 16 L5 and 1 L4 Mars Trojan asteroids. We then combine sample variance analysis with statistical hypothesis testing to identify clusters in the distribution of orbits and assess their significance.
Results. We identify two small clusterings significant at 95% confidence of three H=20-21 asteroids each and investigate their likely origin. One of the clusters is probably the result of rotational breakup of a Eureka family asteroid ~10^8 yr ago. The significantly higher tadpole libration width of asteroids in the other cluster is more consistent with an origin as impact ejecta from Eureka itself and on a timescale comparable to the ~1 Gyr age of its family. We further confirm the previously reported correlations in Eureka family orbital distribution attributed to the long-term action of radiation-driven forces and torques on the asteroids. | [
{
"version": "v1",
"created": "Sat, 22 Feb 2025 14:04:21 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Christou",
"Apostolos A.",
""
],
[
"Georgakarakos",
"Nikos",
""
],
[
"Cuk",
"Matija",
""
],
[
"Dell'Oro",
"Aldo",
""
],
[
"Marshall-Lee",
"Andrew",
""
],
[
"Humpage",
"Alice",
""
]
] |
2502.17349 | Minyeong Hwang | Minyeong Hwang, Ziseok Lee, Kwang-Soo Kim, Kyungsu Kim, Eunho Yang | HybridLinker: Topology-Guided Posterior Sampling for Enhanced Diversity and Validity in 3D Molecular Linker Generation | null | null | null | null | physics.chem-ph cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Linker generation is critical in drug discovery applications such as lead optimization and PROTAC design, where molecular fragments are assembled into diverse drug candidates via molecular linker. Existing methods fall into point cloud-free and point cloud-aware categories based on their use of fragments' 3D poses alongside their topologies in sampling the linker's topology. Point cloud-free models prioritize sample diversity but suffer from lower validity due to overlooking fragments' spatial constraints, while point cloud-aware models ensure higher validity but restrict diversity by enforcing strict spatial constraints. To overcome these trade-offs without additional training, we propose HybridLinker, a framework that enhances point cloud-aware inference by providing diverse bonding topologies from a pretrained point cloud-free model as guidance. At its core, we propose LinkerDPS, the first diffusion posterior sampling (DPS) method operating across point cloud-free and point cloud-aware spaces, bridging molecular topology with 3D point clouds via an energy-inspired function. By transferring the diverse sampling distribution of point cloud-free models into the point cloud-aware distribution, HybridLinker significantly surpasses baselines, improving both validity and diversity in foundational molecular design and applied drug optimization tasks, establishing a new DPS framework in the molecular domains beyond imaging. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 17:23:40 GMT"
},
{
"version": "v2",
"created": "Sat, 01 Mar 2025 16:25:36 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 07:43:18 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Hwang",
"Minyeong",
""
],
[
"Lee",
"Ziseok",
""
],
[
"Kim",
"Kwang-Soo",
""
],
[
"Kim",
"Kyungsu",
""
],
[
"Yang",
"Eunho",
""
]
] |
2502.17356 | Rosie Zhao | Rosie Zhao, Tian Qin, David Alvarez-Melis, Sham Kakade, Naomi Saphra | Distributional Scaling for Emergent Capabilities | 18 pages | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper explores the nature of sudden breakthroughs in language model performance at scale, which stand in contrast to smooth improvements governed by scaling laws. While advocates of "emergence" view breakthroughs as unlocked capabilities, others attribute them to thresholding effects on noncontinuous metrics. We propose that breakthroughs are instead driven by continuous changes in the probability distribution of training outcomes when performance is bimodally distributed across random seeds. In synthetic length generalization tasks, we show that different random seeds can produce either highly linear or emergent scaling trends. We reveal that sharp breakthroughs in metrics are produced by underlying continuous changes in their distribution across seeds. Furthermore, we provide a case study of inverse scaling. We validate our distributional scaling framework on realistic settings by measuring MMLU performance in LM populations. These insights emphasize the role of random variation in the effect of scale on LM capabilities. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 17:34:45 GMT"
},
{
"version": "v2",
"created": "Fri, 28 Feb 2025 01:52:47 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 17:06:39 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhao",
"Rosie",
""
],
[
"Qin",
"Tian",
""
],
[
"Alvarez-Melis",
"David",
""
],
[
"Kakade",
"Sham",
""
],
[
"Saphra",
"Naomi",
""
]
] |
2502.17442 | Xiaoqing Zhang | Xiaoqing Zhang, Yuhan Liu, Flood Sung, Xiuying Chen, Shuo Shang, and Rui Yan | Thinking Before Running! Efficient Code Generation with Thorough Exploration and Optimal Refinement | 14 pages, 10 figures, 8 tables | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Code generation is crucial in software engineering for automating the coding process efficiently. While test-time computation methods show promise, they suffer from high latency due to multiple computation rounds. To overcome this, we introduce \textbf{ThinkCoder}, a framework that combines thorough exploration with optimal refinement. The exploration phase diversifies the solution space by searching for potential solutions, followed by a refinement phase that enhances precision. This approach allows us to select the best solution through careful consideration before taking action, avoiding excessive trial and error. To further minimize test-time computation overhead, we introduce preference-driven optimization with Reinforced Self-Training (ReST), which uses exploration trajectories from ThinkCoder to guide LLM's evolution. This approach enhances LLM's exploration efficiency via preference learning, cutting costs while maintaining accuracy. ThinkCoder boosts the performance with a single LLM, excelling on benchmarks like HumanEval and MBPP. Compared to SOTA models, it improves Pass@1 by 3.0\% over MapCoder with just 6.4\% of the computation cost. Against AgentCoder, ThinkCoder achieves a 0.5\% higher Pass@1 after 2 rounds, outperforming AgentCoder's 5 rounds. Additionally, ReST with success trajectories enhances efficiency, allowing models like LLaMA2-7B to achieve competitive results using only 20\% of the computational resources. These results highlight the framework's effectiveness and scalability. | [
{
"version": "v1",
"created": "Mon, 30 Dec 2024 07:02:15 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 11:36:00 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhang",
"Xiaoqing",
""
],
[
"Liu",
"Yuhan",
""
],
[
"Sung",
"Flood",
""
],
[
"Chen",
"Xiuying",
""
],
[
"Shang",
"Shuo",
""
],
[
"Yan",
"Rui",
""
]
] |
2502.17477 | Niels Lorenzen | Niels R. Lorenzen, Poul J. Jennum, Emmanuel Mignot, Andreas Brink-Kjaer | Frequency-Aware Masked Autoencoders for Human Activity Recognition using Accelerometers | 7 pages, 4 figures, accepted in 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) | null | null | null | eess.SP cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Wearable accelerometers are widely used for continuous monitoring of physical activity. Supervised machine learning and deep learning algorithms have long been used to extract meaningful activity information from raw accelerometry data, but progress has been hampered by the limited amount of labeled data that is publicly available. Exploiting large unlabeled datasets using self-supervised pretraining is a relatively new and underexplored approach in the field of human activity recognition (HAR). We used a time-series transformer masked autoencoder (MAE) approach to self-supervised pretraining and propose two novel spectrogram-based loss functions: the log-scale meanmagnitude (LMM) and log-scale magnitude variance (LMV) losses. We compared these losses with the mean squared error (MSE) loss for MAE training. We leveraged the large unlabeled UK Biobank accelerometry dataset (n = 109k) for pretraining and evaluated downstream HAR performance using a linear classifier in a smaller labelled dataset. We found that pretraining with the LMM loss improved performance compared to an MAE pretrained with the MSE loss, with 12.7% increase in subject-wise F1 score when using linear probing. Compared with a state-of-the-art ResNet-based HAR model, our LMM-pretrained transformer models performed better (+9.8% F1) with linear probing and comparably when fine-tuned using an LSTM classifier. The addition of the LMV to the LMM loss decreased performance compared to the LMM loss alone. These findings establish the LMM loss as a robust and effective method for pretraining MAE models on accelerometer data for HAR and show the potential of pretraining sequence-based models for free-living HAR. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 14:57:51 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 08:46:37 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Lorenzen",
"Niels R.",
""
],
[
"Jennum",
"Poul J.",
""
],
[
"Mignot",
"Emmanuel",
""
],
[
"Brink-Kjaer",
"Andreas",
""
]
] |
2502.17543 | Fahim Tajwar | Fahim Tajwar, Yiding Jiang, Abitha Thankaraj, Sumaita Sadia Rahman, J Zico Kolter, Jeff Schneider, Ruslan Salakhutdinov | Training a Generally Curious Agent | ICML 2025. Project Website: https://paprika-llm.github.io | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Efficient exploration is essential for intelligent systems interacting with their environment, but existing language models often fall short in scenarios that require strategic information gathering. In this paper, we present Paprika, a fine-tuning approach that enables language models to develop general decision-making capabilities that are not confined to particular environments. By training on synthetic interaction data from different tasks that require diverse strategies, Paprika teaches models to explore and adapt their behavior on a new task based on environment feedback in-context without more gradient updates. Experimental results show that models fine-tuned with Paprika can effectively transfer their learned decision-making capabilities to entirely unseen tasks without additional training. Unlike traditional training, our approach's primary bottleneck lies in sampling useful interaction data instead of model updates. To improve sample efficiency, we propose a curriculum learning strategy that prioritizes sampling trajectories from tasks with high learning potential. These results suggest a promising path towards AI systems that can autonomously solve novel sequential decision-making problems that require interactions with the external world. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 18:56:58 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Mar 2025 06:53:52 GMT"
},
{
"version": "v3",
"created": "Mon, 26 May 2025 23:29:19 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Tajwar",
"Fahim",
""
],
[
"Jiang",
"Yiding",
""
],
[
"Thankaraj",
"Abitha",
""
],
[
"Rahman",
"Sumaita Sadia",
""
],
[
"Kolter",
"J Zico",
""
],
[
"Schneider",
"Jeff",
""
],
[
"Salakhutdinov",
"Ruslan",
""
]
] |
2502.17683 | Karl Pierce | Karl Pierce, Miguel Morales | Using Matrix-Free Tensor-Network Optimizations to Construct a Reduced-Scaling and Robust Second-Order M{\o}ller-Plesset Theory | null | null | null | null | physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the efficient combination of the canonical polyadic decomposition (CPD) and tensor hyper-contraction (THC) approaches. We first present a novel low-cost CPD solver which leverages a precomputed THC factorization of an order-$4$ tensor to efficiently optimize the order-$4$ CPD with $\mathcal{O}(NR^2)$ scaling. With the matrix-free THC-based optimization strategy in hand we can: efficiently generate CPD factorizations of the order-4 two-electron integral tensors; and develop novel electronic structure methods which take advantage of both the THC and CPD approximations. Next, we investigate the application of a combined CPD and THC approximation of the Laplace transform (LT) second-order M{\o}ller-Plesset (MP2) method. We exploit the ability to switch efficiently between the THC and CPD factorizations of the two electron integrals to reduce the computational complexity of the LT MP2 method while preserving the accuracy of the approach. Furthermore we take advantage of the robust fitting approximation to eliminate leading order error in the CPD approximated tensor networks. Finally, we show that modest values of THC and CPD rank preserve the accuracy of the LT MP2 method and that this CPD+THC LT MP2 strategy realizes a performance advantage over canonical LT MP2 in both computational wall-times and memory resource requirements. | [
{
"version": "v1",
"created": "Mon, 24 Feb 2025 22:06:02 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 00:46:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Pierce",
"Karl",
""
],
[
"Morales",
"Miguel",
""
]
] |
2502.17817 | Md Kowsher | Md Kowsher, Nusrat Jahan Prottasha, Prakash Bhat, Chun-Nam Yu, Mojtaba Soltanalian, Ivan Garibay, Ozlem Garibay, Chen Chen, Niloofar Yousefi | Predicting Through Generation: Why Generation Is Better for Prediction | ACL Accepted paper | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | This paper argues that generating output tokens is more effective than using pooled representations for prediction tasks because token-level generation retains more mutual information. Since LLMs are trained on massive text corpora using next-token prediction, generation aligns naturally with their learned behavior. Using the Data Processing Inequality (DPI), we provide both theoretical and empirical evidence supporting this claim. However, autoregressive models face two key challenges when used for prediction: (1) exposure bias, where the model sees ground truth tokens during training but relies on its own predictions during inference, leading to errors, and (2) format mismatch, where discrete tokens do not always align with the tasks required output structure. To address these challenges, we introduce PredGen(Predicting Through Generating), an end to end framework that (i) uses scheduled sampling to reduce exposure bias, and (ii) introduces a task adapter to convert the generated tokens into structured outputs. Additionally, we introduce Writer-Director Alignment Loss (WDAL), which ensures consistency between token generation and final task predictions, improving both text coherence and numerical accuracy. We evaluate PredGen on multiple classification and regression benchmarks. Our results show that PredGen consistently outperforms standard baselines, demonstrating its effectiveness in structured prediction tasks. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 03:48:19 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 22:30:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kowsher",
"Md",
""
],
[
"Prottasha",
"Nusrat Jahan",
""
],
[
"Bhat",
"Prakash",
""
],
[
"Yu",
"Chun-Nam",
""
],
[
"Soltanalian",
"Mojtaba",
""
],
[
"Garibay",
"Ivan",
""
],
[
"Garibay",
"Ozlem",
""
],
[
"Chen",
"Chen",
""
],
[
"Yousefi",
"Niloofar",
""
]
] |
2502.17906 | Yuki Sato | Yuki Sato and Kiyoshi Kanazawa | Why do financial prices exhibit Brownian motion despite predictable order flow? | Main: 7 pages, 4 figures. SI: 6 pages, 3 figures. Minor bugs in simulation codes are fixed | null | null | null | q-fin.TR cond-mat.stat-mech econ.GN q-fin.EC q-fin.MF q-fin.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In financial market microstructure, there are two enigmatic empirical laws: (i) the market-order flow has predictable persistence due to metaorder splitters by institutional investors, well formulated as the Lillo-Mike-Farmer model. However, this phenomenon seems paradoxical given the diffusive and unpredictable price dynamics; (ii) the price impact $I(Q)$ of a large metaorder $Q$ follows the square-root law, $I(Q)\propto \sqrt{Q}$. Here we theoretically reveal why price dynamics follows Brownian motion despite predictable order flow by unifying these enigmas. We generalize the Lillo-Mike-Farmer model to nonlinear price-impact dynamics, which is mapped to an exactly solvable L\'evy-walk model. Our exact solution shows that the price dynamics remains diffusive under the square-root law, even under persistent order flow. This work illustrates the crucial role of the square-root law in mitigating large price movements by large metaorders, thereby leading to the Brownian price dynamics, consistently with the efficient market hypothesis over long timescales. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 07:12:03 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Mar 2025 11:49:37 GMT"
},
{
"version": "v3",
"created": "Tue, 06 May 2025 08:45:34 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 09:32:30 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Sato",
"Yuki",
""
],
[
"Kanazawa",
"Kiyoshi",
""
]
] |
2502.18001 | Xinghao Chen | Xinghao Chen, Zhijing Sun, Wenjin Guo, Miaoran Zhang, Yanjun Chen, Yirong Sun, Hui Su, Yijie Pan, Dietrich Klakow, Wenjie Li, Xiaoyu Shen | Unveiling the Key Factors for Distilling Chain-of-Thought Reasoning | ACL 2025 Findings | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Large Language Models (LLMs) excel in reasoning tasks through Chain-of-Thought (CoT) prompting. However, CoT prompting greatly increases computational demands, which has prompted growing interest in distilling CoT capabilities into Small Language Models (SLMs). This study systematically examines the factors influencing CoT distillation, including the choice of granularity, format and teacher model. Through experiments involving four teacher models and seven student models across seven mathematical and commonsense reasoning datasets, we uncover three key findings: (1) Unlike LLMs, SLMs exhibit a non-monotonic relationship with granularity, with stronger models benefiting from finer-grained reasoning and weaker models performing better with simpler CoT supervision; (2) CoT format significantly impacts LLMs but has minimal effect on SLMs, likely due to their reliance on supervised fine-tuning rather than pretraining preferences; (3) Stronger teacher models do NOT always produce better student models, as diversity and complexity in CoT supervision can outweigh accuracy alone. These findings emphasize the need to tailor CoT strategies to specific student model, offering actionable insights for optimizing CoT distillation in SLMs. The code and datasets are available at https://github.com/EIT-NLP/Distilling-CoT-Reasoning. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 09:08:45 GMT"
},
{
"version": "v2",
"created": "Sun, 25 May 2025 08:05:57 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:59:22 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chen",
"Xinghao",
""
],
[
"Sun",
"Zhijing",
""
],
[
"Guo",
"Wenjin",
""
],
[
"Zhang",
"Miaoran",
""
],
[
"Chen",
"Yanjun",
""
],
[
"Sun",
"Yirong",
""
],
[
"Su",
"Hui",
""
],
[
"Pan",
"Yijie",
""
],
[
"Klakow",
"Dietrich",
""
],
[
"Li",
"Wenjie",
""
],
[
"Shen",
"Xiaoyu",
""
]
] |
2502.18351 | Konrad Steinvall | Konrad Steinvall, Louis Richard, T\"unde F\"ul\"op, Lise Hanebring, Istv\'an Pusztai | Eulerian and Lagrangian electron energization during magnetic reconnection | 18 pages, 7 figures | null | 10.1017/S0022377825100408 | null | physics.plasm-ph physics.space-ph | http://creativecommons.org/licenses/by/4.0/ | Electron energization by magnetic reconnection has historically been studied in the Lagrangian guiding-center framework. Insights from such studies include that Fermi acceleration in magnetic islands can accelerate electrons to high energies. An alternative Eulerian fluid formulation of electron energization was recently used to study electron energization during magnetic reconnection in the absence of magnetic islands. Here, we use particle-in-cell simulations to compare the Eulerian and Lagrangian models of electron energization in a setup where reconnection leads to magnetic island formation. We find the largest energization at the edges of magnetic islands. There, energization related to the diamagnetic drift dominates in the Eulerian model, while the Fermi related term dominates in the Lagrangian model. The models predict significantly different energization rates locally. A better agreement is found after integrating over the simulation domain. We show that strong magnetic curvature can break the magnetic moment conservation assumed by the Lagrangian model, leading to erroneous results. The Eulerian fluid model is a complete fluid description and accurately models bulk energization. However, local measurements of its constituent energization terms need not reflect locations where plasma is heated or accelerated. The Lagrangian guiding center model can accurately describe the energization of particles, but it cannot describe the evolution of the fluid energy. We conclude that while both models can be valid, they describe two fundamentally different quantities, and care should be taken when choosing which model to use. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 16:41:42 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Apr 2025 12:01:03 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Steinvall",
"Konrad",
""
],
[
"Richard",
"Louis",
""
],
[
"Fülöp",
"Tünde",
""
],
[
"Hanebring",
"Lise",
""
],
[
"Pusztai",
"István",
""
]
] |
2502.18615 | Georgios Kamaras | Georgios Kamaras and Subramanian Ramamoorthy | A Distributional Treatment of Real2Sim2Real for Object-Centric Agent Adaptation in Vision-Driven Deformable Linear Object Manipulation | null | null | null | null | cs.RO cs.LG | http://creativecommons.org/licenses/by/4.0/ | We present an integrated (or end-to-end) framework for the Real2Sim2Real problem of manipulating deformable linear objects (DLOs) based on visual perception. Working with a parameterised set of DLOs, we use likelihood-free inference (LFI) to compute the posterior distributions for the physical parameters using which we can approximately simulate the behaviour of each specific DLO. We use these posteriors for domain randomisation while training, in simulation, object-specific visuomotor policies (i.e. assuming only visual and proprioceptive sensory) for a DLO reaching task, using model-free reinforcement learning. We demonstrate the utility of this approach by deploying sim-trained DLO manipulation policies in the real world in a zero-shot manner, i.e. without any further fine-tuning. In this context, we evaluate the capacity of a prominent LFI method to perform fine classification over the parametric set of DLOs, using only visual and proprioceptive data obtained in a dynamic manipulation trajectory. We then study the implications of the resulting domain distributions in sim-based policy learning and real-world performance. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 20:01:06 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 12:56:53 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kamaras",
"Georgios",
""
],
[
"Ramamoorthy",
"Subramanian",
""
]
] |
2502.18624 | Nikolaos Fytas G. | Mauro Bisson, Massimo Bernaschi, Massimiliano Fatica, Nikolaos G. Fytas, Isidoro Gonz\'alez-Adalid Pemart\'in, V\'ictor Mart\'in-Mayor, Alexandros Vasilopoulos | Massive-Scale Simulations of 2D Ising and Blume-Capel Models on Rack-Scale Multi-GPU Systems | 18 pages, 4 tables, 5 figures, version accepted for publication in Computer Physics Communications | null | null | null | physics.comp-ph cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present high-performance implementations of the two-dimensional Ising and Blume-Capel models for large-scale, multi-GPU simulations. Our approach takes full advantage of the NVIDIA GB200 NVL72 system, which features up to $72$ GPUs interconnected via high-bandwidth NVLink, enabling direct GPU-to-GPU memory access across multiple nodes. By utilizing Fabric Memory and an optimized Monte Carlo kernel for the Ising model, our implementation supports simulations of systems with linear sizes up to $L=2^{23}$, corresponding to approximately $70$ trillion spins. This allows for a peak processing rate of nearly $1.15 \times 10^5$ lattice updates per nanosecond-setting a new performance benchmark for Ising model simulations. Additionally, we introduce a custom protocol for computing correlation functions, which strikes an optimal balance between computational efficiency and statistical accuracy. This protocol enables large-scale simulations without incurring prohibitive runtime costs. Benchmark results show near-perfect strong and weak scaling up to $64$ GPUs, demonstrating the effectiveness of our approach for large-scale statistical physics simulations. | [
{
"version": "v1",
"created": "Tue, 25 Feb 2025 20:18:08 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 04:14:58 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bisson",
"Mauro",
""
],
[
"Bernaschi",
"Massimo",
""
],
[
"Fatica",
"Massimiliano",
""
],
[
"Fytas",
"Nikolaos G.",
""
],
[
"Pemartín",
"Isidoro González-Adalid",
""
],
[
"Martín-Mayor",
"Víctor",
""
],
[
"Vasilopoulos",
"Alexandros",
""
]
] |
2502.18869 | Mrinal Kumar Das | Ankita Kakoti and Mrinal Kumar Das | keV sterile neutrino dark matter and associated phenomenology in doublet left-right symmetric model | 31 pages, 13 figures. arXiv admin note: text overlap with arXiv:2312.00822, arXiv:2406.02944 | null | null | null | hep-ph | http://creativecommons.org/licenses/by/4.0/ | Considering the fact that no Standard Model (SM) particle can account as a suitable dark matter candidate in the relevant studies, several Beyond Standard Model (BSM) frameworks incorporate within itself some extra fields depending upon the model taken into account. Left-Right Symmetric Model (LRSM) is one such BSM framework which can successfully explain the origin of neutrino masses and also give an elaborative elucidation of the associated phenomenology like Neutrinoless Double Beta Decay $(0\nu\beta\beta)$, Lepton Flavor Violation (LFV), baryogenesis via leptogenesis etc. However, in spite of the presence of several new fields within the model, none of them can stand as a suitable dark matter candidate. This compels us to extend LRSM with extra scalars or fermions depending upon the type of study to be taken into account. In the current work, we extend LRSM with a sterile fermion per generation and the lightest of which acts as the Dark Matter (DM) candidate. The realization of the model has been done using modular symmetry, where we have used modular group $\Gamma(3)$ of weight 2 which is isomorphic to non-abelian discrete symmetry group $A_{4}$. We have then calculated the relic abundance as well the decay rate of the corresponding DM candidate and also studied $0\nu\beta\beta$ and LFV within the model and the results have been discussed in detail within the manuscript. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 06:26:57 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:01:26 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kakoti",
"Ankita",
""
],
[
"Das",
"Mrinal Kumar",
""
]
] |
2502.18874 | Kaishuai Xu | Kaishuai Xu, Tiezheng Yu, Wenjun Hou, Yi Cheng, Liangyou Li, Xin Jiang, Lifeng Shang, Qun Liu, Wenjie Li | Learning to Align Multi-Faceted Evaluation: A Unified and Robust Framework | accepted as ACL 2025 findings | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) are being used more and more extensively for automated evaluation in various scenarios. Previous studies have attempted to fine-tune open-source LLMs to replicate the evaluation explanations and judgments of powerful proprietary models, such as GPT-4. However, these methods are largely limited to text-based analyses under predefined general criteria, resulting in reduced adaptability for unseen instructions and demonstrating instability in evaluating adherence to quantitative and structural constraints. To address these limitations, we propose a novel evaluation framework, ARJudge, that adaptively formulates evaluation criteria and synthesizes both text-based and code-driven analyses to evaluate LLM responses. ARJudge consists of two components: a fine-tuned Analyzer that generates multi-faceted evaluation analyses and a tuning-free Refiner that combines and refines all analyses to make the final judgment. We construct a Composite Analysis Corpus that integrates tasks for evaluation criteria generation alongside text-based and code-driven analysis generation to train the Analyzer. Our results demonstrate that ARJudge outperforms existing fine-tuned evaluators in effectiveness and robustness. Furthermore, it demonstrates the importance of multi-faceted evaluation and code-driven analyses in enhancing evaluation capabilities. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 06:31:45 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Mar 2025 07:13:12 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 03:32:22 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Xu",
"Kaishuai",
""
],
[
"Yu",
"Tiezheng",
""
],
[
"Hou",
"Wenjun",
""
],
[
"Cheng",
"Yi",
""
],
[
"Li",
"Liangyou",
""
],
[
"Jiang",
"Xin",
""
],
[
"Shang",
"Lifeng",
""
],
[
"Liu",
"Qun",
""
],
[
"Li",
"Wenjie",
""
]
] |
2502.19130 | Lars Kaesberg | Lars Benedikt Kaesberg, Jonas Becker, Jan Philip Wahle, Terry Ruas, Bela Gipp | Voting or Consensus? Decision-Making in Multi-Agent Debate | null | null | null | null | cs.MA cs.AI cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Much of the success of multi-agent debates depends on carefully choosing the right parameters. The decision-making protocol stands out as it can highly impact final model answers, depending on how decisions are reached. Systematic comparison of decision protocols is difficult because many studies alter multiple discussion parameters beyond the protocol. So far, it has been largely unknown how decision-making influences different tasks. This work systematically evaluates the impact of seven decision protocols (e.g., majority voting, unanimity consensus). We change only one variable at a time - the decision protocol - to analyze how different methods affect the collaboration between agents and measure differences in knowledge and reasoning tasks. Our results show that voting protocols improve performance by 13.2% in reasoning tasks and consensus protocols by 2.8% in knowledge tasks compared to other decision protocols. Increasing the number of agents improves performance, while more discussion rounds before voting reduce it. To improve decision-making by increasing answer diversity, we propose two new methods, All-Agents Drafting (AAD) and Collective Improvement (CI). Our methods improve task performance by up to 3.3% with AAD and up to 7.4% with CI. This work demonstrates the importance of decision-making in multi-agent debates beyond scaling. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 13:39:18 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 13:52:26 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kaesberg",
"Lars Benedikt",
""
],
[
"Becker",
"Jonas",
""
],
[
"Wahle",
"Jan Philip",
""
],
[
"Ruas",
"Terry",
""
],
[
"Gipp",
"Bela",
""
]
] |
2502.19249 | Michael Hu | Michael Y. Hu, Jackson Petty, Chuan Shi, William Merrill, Tal Linzen | Between Circuits and Chomsky: Pre-pretraining on Formal Languages Imparts Linguistic Biases | ACL 2025 Camera Ready | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Pretraining language models on formal language can improve their acquisition of natural language. Which features of the formal language impart an inductive bias that leads to effective transfer? Drawing on insights from linguistics and complexity theory, we hypothesize that effective transfer occurs when two conditions are met: the formal language should capture the dependency structures present in natural language, and it should remain within the computational limitations of the model architecture. We experiment with pre-pretraining (training on formal language before natural languages) on transformers and find that formal languages capturing hierarchical dependencies indeed enable language models to achieve lower loss on natural language and better linguistic generalization compared to other formal languages. We also find modest support for the hypothesis that the formal language should fall within the computational limitations of the architecture. Strikingly, pre-pretraining reduces loss more efficiently than training on a matched amount of natural language. For a 1B-parameter language model trained on roughly 1.6B tokens of natural language, pre-pretraining achieves the same loss and better linguistic generalization with a 33% smaller token budget. Finally, we also give mechanistic evidence of transfer from formal to natural language: attention heads acquired during pre-pretraining remain crucial for the model's performance on syntactic evaluations. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 15:55:55 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 04:23:34 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Hu",
"Michael Y.",
""
],
[
"Petty",
"Jackson",
""
],
[
"Shi",
"Chuan",
""
],
[
"Merrill",
"William",
""
],
[
"Linzen",
"Tal",
""
]
] |
2502.19311 | Christoph Benzm\"uller | Christoph Benzm\"uller | Faithful Logic Embeddings in HOL -- Deep and Shallow | Preprint of paper accepted for CADE 2025; 24 pages, 11 figures | null | null | null | cs.LO cs.AI math.LO | http://creativecommons.org/licenses/by/4.0/ | Deep and shallow embeddings of non-classical logics in classical higher-order logic have been explored, implemented, and used in various reasoning tools in recent years. This paper presents a method for the simultaneous deployment of deep and shallow embeddings of various degrees in classical higher-order logic. This enables flexible, interactive and automated theorem proving and counterexample finding at meta and object level, as well as automated faithfulness proofs between these logic embeddings. The method is beneficial for logic education, research and application and is illustrated here using a simple propositional modal logic. However, this approach is conceptual in nature and not limited to this simple logical context. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 17:08:07 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 14:41:21 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Benzmüller",
"Christoph",
""
]
] |
2502.19459 | Yu Liu | Yu Liu, Baoxiong Jia, Ruijie Lu, Junfeng Ni, Song-Chun Zhu, Siyuan Huang | ArtGS: Building Interactable Replicas of Complex Articulated Objects via Gaussian Splatting | null | null | null | null | cs.CV cs.GR cs.LG cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Building articulated objects is a key challenge in computer vision. Existing methods often fail to effectively integrate information across different object states, limiting the accuracy of part-mesh reconstruction and part dynamics modeling, particularly for complex multi-part articulated objects. We introduce ArtGS, a novel approach that leverages 3D Gaussians as a flexible and efficient representation to address these issues. Our method incorporates canonical Gaussians with coarse-to-fine initialization and updates for aligning articulated part information across different object states, and employs a skinning-inspired part dynamics modeling module to improve both part-mesh reconstruction and articulation learning. Extensive experiments on both synthetic and real-world datasets, including a new benchmark for complex multi-part objects, demonstrate that ArtGS achieves state-of-the-art performance in joint parameter estimation and part mesh reconstruction. Our approach significantly improves reconstruction quality and efficiency, especially for multi-part articulated objects. Additionally, we provide comprehensive analyses of our design choices, validating the effectiveness of each component to highlight potential areas for future improvement. Our work is made publicly available at: https://articulate-gs.github.io. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 10:25:32 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Mar 2025 08:43:16 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Liu",
"Yu",
""
],
[
"Jia",
"Baoxiong",
""
],
[
"Lu",
"Ruijie",
""
],
[
"Ni",
"Junfeng",
""
],
[
"Zhu",
"Song-Chun",
""
],
[
"Huang",
"Siyuan",
""
]
] |
2502.20007 | Muhammed Deniz | S. Karadag, M. Deniz, S. Karmakar, M. K. Singh, M. Demirci, Greeshma C., H. B. Li, S. T. Lin, M. F. Mustamin, V. Sharma, L. Singh, M. K. Singh, V. Singh, H. T. Wong | Constraints on New Physics with Light Mediators and Generalized Neutrino Interactions via Coherent Elastic Neutrino Nucleus Scattering | 28 pages, 20 figures | null | null | null | hep-ex hep-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We investigate new physics effects on the framework of Coherent Elastic Neutrino Nucleus Scattering (CE$\nu$NS) from Non-Standard Interactions (NSI) of neutrinos and generalized neutrino interactions (GNI) as well as the possibility of light mediators from simplified model which includes all possible Lorentz-invariant interactions of vector (V), axialvector (A), scalar (S), pseudoscalar (P), and tensor (T) type. The constraints and the allowed regions at $90\%$ C.L. for masses and couplings in each new physics scenario are derived from the analysis of TEXONO data, including two datasets from 2016 with a high purity $n$-type point-contact (nPC) germanium (Ge) detector and from 2024 with an advanced $p$-type point-contact (pPC) Ge detector. The results are presented in comparison with the current data of COHERENT with CsI and LAr detectors to highlight the complementarity of reactor- and accelerator-based neutrino experiments. | [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 11:38:32 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:15:58 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Karadag",
"S.",
""
],
[
"Deniz",
"M.",
""
],
[
"Karmakar",
"S.",
""
],
[
"Singh",
"M. K.",
""
],
[
"Demirci",
"M.",
""
],
[
"C.",
"Greeshma",
""
],
[
"Li",
"H. B.",
""
],
[
"Lin",
"S. T.",
""
],
[
"Mustamin",
"M. F.",
""
],
[
"Sharma",
"V.",
""
],
[
"Singh",
"L.",
""
],
[
"Singh",
"M. K.",
""
],
[
"Singh",
"V.",
""
],
[
"Wong",
"H. T.",
""
]
] |
2502.20074 | Jan Verlage | Jan Verlage and Peter Kratzer | Transmission through multiple Mott insulator - semiconductor wells | null | J. Verlage and P. Kratzer, Phys. Rev. B 111, 195158 (2025) | 10.1103/PhysRevB.111.195158 | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Weakly and strongly interacting quantum many-body systems, namely semiconductors and Mott insulators, are combined into a layered heterostructure. Via the hierarchy of correlations, we derive and match the propagating quasi-particle solutions in the different regions and calculate the transmission coefficients through these layered structures. As a proof of principle, we find the well known transmission bands of a semiconductor heterostructure. Extending this idea to semiconductor and Mott insulator structures we calculate the transmittance and the resonance energies. Within a phase accumulation model we find analytical expressions for the scattering phase shift. Lastly, we find transmission curves with skewness for structures with applied voltage. | [
{
"version": "v1",
"created": "Thu, 27 Feb 2025 13:31:14 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:52:00 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Verlage",
"Jan",
""
],
[
"Kratzer",
"Peter",
""
]
] |
2502.20411 | Saeed Reza Kheradpisheh | Mohammadnavid Ghader, Saeed Reza Kheradpisheh, Bahar Farahani, Mahmood Fazlali | Backpropagation-free Spiking Neural Networks with the Forward-Forward Algorithm | null | null | null | null | cs.NE cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Spiking Neural Networks (SNNs) offer a biologically inspired computational paradigm that emulates neuronal activity through discrete spike-based processing. Despite their advantages, training SNNs with traditional backpropagation (BP) remains challenging due to computational inefficiencies and a lack of biological plausibility. This study explores the Forward-Forward (FF) algorithm as an alternative learning framework for SNNs. Unlike backpropagation, which relies on forward and backward passes, the FF algorithm employs two forward passes, enabling layer-wise localized learning, enhanced computational efficiency, and improved compatibility with neuromorphic hardware. We introduce an FF-based SNN training framework and evaluate its performance across both non-spiking (MNIST, Fashion-MNIST, Kuzushiji-MNIST) and spiking (Neuro-MNIST, SHD) datasets. Experimental results demonstrate that our model surpasses existing FF-based SNNs on evaluated static datasets with a much lighter architecture while achieving accuracy comparable to state-of-the-art backpropagation-trained SNNs. On more complex spiking tasks such as SHD, our approach outperforms other SNN models and remains competitive with leading backpropagation-trained SNNs. These findings highlight the FF algorithm's potential to advance SNN training methodologies by addressing some key limitations of backpropagation. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 12:44:26 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 12:47:13 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ghader",
"Mohammadnavid",
""
],
[
"Kheradpisheh",
"Saeed Reza",
""
],
[
"Farahani",
"Bahar",
""
],
[
"Fazlali",
"Mahmood",
""
]
] |
2502.20606 | Joao Marcos Correia Marques | Joao Marcos Correia Marques, Nils Dengler, Tobias Zaenker, Jesper Mucke, Shenlong Wang, Maren Bennewitz and Kris Hauser | Map Space Belief Prediction for Manipulation-Enhanced Mapping | 14 pages, 10 figures; To appear at RSS 2025 | null | null | null | cs.RO cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Searching for objects in cluttered environments requires selecting efficient viewpoints and manipulation actions to remove occlusions and reduce uncertainty in object locations, shapes, and categories. In this work, we address the problem of manipulation-enhanced semantic mapping, where a robot has to efficiently identify all objects in a cluttered shelf. Although Partially Observable Markov Decision Processes~(POMDPs) are standard for decision-making under uncertainty, representing unstructured interactive worlds remains challenging in this formalism. To tackle this, we define a POMDP whose belief is summarized by a metric-semantic grid map and propose a novel framework that uses neural networks to perform map-space belief updates to reason efficiently and simultaneously about object geometries, locations, categories, occlusions, and manipulation physics. Further, to enable accurate information gain analysis, the learned belief updates should maintain calibrated estimates of uncertainty. Therefore, we propose Calibrated Neural-Accelerated Belief Updates (CNABUs) to learn a belief propagation model that generalizes to novel scenarios and provides confidence-calibrated predictions for unknown areas. Our experiments show that our novel POMDP planner improves map completeness and accuracy over existing methods in challenging simulations and successfully transfers to real-world cluttered shelves in zero-shot fashion. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 00:10:52 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 05:37:31 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Marques",
"Joao Marcos Correia",
""
],
[
"Dengler",
"Nils",
""
],
[
"Zaenker",
"Tobias",
""
],
[
"Mucke",
"Jesper",
""
],
[
"Wang",
"Shenlong",
""
],
[
"Bennewitz",
"Maren",
""
],
[
"Hauser",
"Kris",
""
]
] |
2502.20795 | Kuang-Da Wang | Kuang-Da Wang, Teng-Ruei Chen, Yu Heng Hung, Guo-Xun Ko, Shuoyang Ding, Yueh-Hua Wu, Yu-Chiang Frank Wang, Chao-Han Huck Yang, Wen-Chih Peng, Ping-Chun Hsieh | Plan2Align: Predictive Planning Based Test-Time Preference Alignment for Large Language Models | Preprint. Code will be released at Plan2Align GitHub link: https://github.com/NYCU-RL-Bandits-Lab/Plan2Align | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Aligning Large Language Models with Preference Fine-Tuning is often resource-intensive. Test-time alignment techniques that do not modify the underlying models, such as prompting and guided decodings, offer a lightweight alternative. However, existing test-time alignment methods primarily improve short responses and fail to ensure coherence over extended contexts due to the myopic nature of token-level alignment. Moreover, these methods often incur a slowdown during inference. To address these challenges, we propose Plan2Align, a test-time alignment framework that formulates text generation as a predictive planning problem. Plan2Align adapts Model Predictive Control (MPC) to iteratively refine output by rolling out multiple complete responses and optimizing each segment. To more rigorously evaluate the effectiveness and efficiency, we focus on the more challenging task of long-text generation. Experiments on the long-form response subset of the HH-RLHF dataset and the WMT'24 Discourse-Level Literary Translation demonstrate that Plan2Align significantly enhances the performance of base LLMs. Compared to existing training-time and test-time alignment methods on LLaMA-3.1 8B, Plan2Align achieves comparable or superior results, while also delivering improved inference efficiency relative to prior test-time alignment approaches. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 07:24:33 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 11:26:15 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Wang",
"Kuang-Da",
""
],
[
"Chen",
"Teng-Ruei",
""
],
[
"Hung",
"Yu Heng",
""
],
[
"Ko",
"Guo-Xun",
""
],
[
"Ding",
"Shuoyang",
""
],
[
"Wu",
"Yueh-Hua",
""
],
[
"Wang",
"Yu-Chiang Frank",
""
],
[
"Yang",
"Chao-Han Huck",
""
],
[
"Peng",
"Wen-Chih",
""
],
[
"Hsieh",
"Ping-Chun",
""
]
] |
2502.20859 | Yifan Duan | Yifan Duan, Yihong Tang, Xuefeng Bai, Kehai Chen, Juntao Li, Min Zhang | The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) excel in both closed tasks (including problem-solving, and code generation) and open tasks (including creative writing), yet existing explanations for their capabilities lack connections to real-world human intelligence. To fill this gap, this paper systematically investigates LLM intelligence through the lens of ``human simulation'', addressing three core questions: (1) \textit{How do personality traits affect problem-solving in closed tasks?} (2) \textit{How do traits shape creativity in open tasks?} (3) \textit{How does single-agent performance influence multi-agent collaboration?} By assigning Big Five personality traits to LLM agents and evaluating their performance in single- and multi-agent settings, we reveal that specific traits significantly influence reasoning accuracy (closed tasks) and creative output (open tasks). Furthermore, multi-agent systems exhibit collective intelligence distinct from individual capabilities, driven by distinguishing combinations of personalities. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 09:01:39 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 07:52:28 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Duan",
"Yifan",
""
],
[
"Tang",
"Yihong",
""
],
[
"Bai",
"Xuefeng",
""
],
[
"Chen",
"Kehai",
""
],
[
"Li",
"Juntao",
""
],
[
"Zhang",
"Min",
""
]
] |
2502.20968 | Weixiang Zhao | Weixiang Zhao, Yulin Hu, Yang Deng, Jiahe Guo, Xingyu Sui, Xinyang Han, An Zhang, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu | Beware of Your Po! Measuring and Mitigating AI Safety Risks in Role-Play Fine-Tuning of LLMs | To appear at ACL 2025 (Main) | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Role-playing enables large language models (LLMs) to engage users in immersive and personalized interactions, but it also introduces significant safety risks. Existing role-play fine-tuning techniques improve role adaptability but may degrade safety performance, particularly for villainous characters. In this work, we conduct the first comprehensive assessment of role-play fine-tuning risks by training 95 role-specific LLMs using RoleBench. Our experiments reveal that role-play fine-tuning leads to a noticeable decline in safety performance, with safety risks varying based on character traits. To tackle this challenge, we propose Safety-Aware Role-Play Fine-Tuning (SaRFT), a novel method designed to balance role-playing capabilities and safety. Extensive experiments on LLaMA-3-8B-Instruct, Gemma-2-9B-it, and Qwen2.5-7B-Instruct demonstrate that SaRFT consistently outperforms state-of-the-art baselines under both LoRA and full-parameter fine-tuning settings. Our findings highlight the necessity of role-adaptive safety measures and provide insights into mitigating role-specific safety risks in role-playing LLMs. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 11:31:27 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 06:26:02 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhao",
"Weixiang",
""
],
[
"Hu",
"Yulin",
""
],
[
"Deng",
"Yang",
""
],
[
"Guo",
"Jiahe",
""
],
[
"Sui",
"Xingyu",
""
],
[
"Han",
"Xinyang",
""
],
[
"Zhang",
"An",
""
],
[
"Zhao",
"Yanyan",
""
],
[
"Qin",
"Bing",
""
],
[
"Chua",
"Tat-Seng",
""
],
[
"Liu",
"Ting",
""
]
] |
2502.20991 | Dieter Spreen | Dieter Spreen | Domains, Information Frames, Rough Sets: An Equivalence of Categories | null | null | null | null | cs.LO math.LO | http://creativecommons.org/licenses/by/4.0/ | A generalization of Scott's information systems~\cite{sco82} is presented that captures exactly all continuous domains. The global consistency predicate in Scott's definition is relativized. Now, for every atomic statement, there is a consistency predicate that states which finite sets of statements express information that is consistent with the given statement. The category of information frames is shown to be equivalent to the category of domains. Moreover, the relationship with CF-approximation spaces introduced by Wu and Xu~\cite{wx23} is studied. The corresponding category is also shown to be equivalent with the category of information frames. This research achieves a refinement of the equivalence result of Wu and Xu of the category of CF-approximation spaces with the category of domains. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 12:19:07 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 12:00:09 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Spreen",
"Dieter",
""
]
] |
2502.21033 | Federica Ferrarese | Muhammad Awais, Abu Safyan Ali, Giacomo Dimarco, Federica Ferrarese, Lorenzo Pareschi | A data augmentation strategy for deep neural networks with application to epidemic modelling | null | null | null | null | math.NA cs.LG cs.NA physics.soc-ph q-bio.PE stat.ML | http://creativecommons.org/licenses/by/4.0/ | In this work, we integrate the predictive capabilities of compartmental disease dynamics models with machine learning ability to analyze complex, high-dimensional data and uncover patterns that conventional models may overlook. Specifically, we present a proof of concept demonstrating the application of data-driven methods and deep neural networks to a recently introduced Susceptible-Infected-Recovered type model with social features, including a saturated incidence rate, to improve epidemic prediction and forecasting. Our results show that a robust data augmentation strategy trough suitable data-driven models can improve the reliability of Feed-Forward Neural Networks and Nonlinear Autoregressive Networks, providing a complementary strategy to Physics-Informed Neural Networks, particularly in settings where data augmentation from mechanistic models can enhance learning. This approach enhances the ability to handle nonlinear dynamics and offers scalable, data-driven solutions for epidemic forecasting, prioritizing predictive accuracy over the constraints of physics-based models. Numerical simulations of the lockdown and post-lockdown phase of the COVID-19 epidemic in Italy and Spain validate our methodology. | [
{
"version": "v1",
"created": "Fri, 28 Feb 2025 13:24:49 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 09:02:35 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Awais",
"Muhammad",
""
],
[
"Ali",
"Abu Safyan",
""
],
[
"Dimarco",
"Giacomo",
""
],
[
"Ferrarese",
"Federica",
""
],
[
"Pareschi",
"Lorenzo",
""
]
] |
2503.00429 | Jinyi Yang | Jingyi Yang, Xun Lin, Zitong Yu, Liepiao Zhang, Xin Liu, Hui Li, Xiaochen Yuan, Xiaochun Cao | DADM: Dual Alignment of Domain and Modality for Face Anti-spoofing | 18 pages, 9 figures, Code: https://github.com/yjyddq/DADM | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the availability of diverse sensor modalities (i.e., RGB, Depth, Infrared) and the success of multi-modal learning, multi-modal face anti-spoofing (FAS) has emerged as a prominent research focus. The intuition behind it is that leveraging multiple modalities can uncover more intrinsic spoofing traces. However, this approach presents more risk of misalignment. We identify two main types of misalignment: (1) \textbf{Intra-domain modality misalignment}, where the importance of each modality varies across different attacks. For instance, certain modalities (e.g., Depth) may be non-defensive against specific attacks (e.g., 3D mask), indicating that each modality has unique strengths and weaknesses in countering particular attacks. Consequently, simple fusion strategies may fall short. (2) \textbf{Inter-domain modality misalignment}, where the introduction of additional modalities exacerbates domain shifts, potentially overshadowing the benefits of complementary fusion. To tackle (1), we propose a alignment module between modalities based on mutual information, which adaptively enhances favorable modalities while suppressing unfavorable ones. To address (2), we employ a dual alignment optimization method that aligns both sub-domain hyperplanes and modality angle margins, thereby mitigating domain gaps. Our method, dubbed \textbf{D}ual \textbf{A}lignment of \textbf{D}omain and \textbf{M}odality (DADM), achieves state-of-the-art performance in extensive experiments across four challenging protocols demonstrating its robustness in multi-modal domain generalization scenarios. The codes will be released soon. | [
{
"version": "v1",
"created": "Sat, 01 Mar 2025 10:12:00 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:28:52 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Yang",
"Jingyi",
""
],
[
"Lin",
"Xun",
""
],
[
"Yu",
"Zitong",
""
],
[
"Zhang",
"Liepiao",
""
],
[
"Liu",
"Xin",
""
],
[
"Li",
"Hui",
""
],
[
"Yuan",
"Xiaochen",
""
],
[
"Cao",
"Xiaochun",
""
]
] |
2503.01885 | Luise Ge | Luise Ge, Michael Lanier, Anindya Sarkar, Bengisu Guresti, Chongjie Zhang, Yevgeniy Vorobeychik | Learning Policy Committees for Effective Personalization in MDPs with Diverse Tasks | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Many dynamic decision problems, such as robotic control, involve a series of tasks, many of which are unknown at training time. Typical approaches for these problems, such as multi-task and meta reinforcement learning, do not generalize well when the tasks are diverse. On the other hand, approaches that aim to tackle task diversity, such as using task embedding as policy context and task clustering, typically lack performance guarantees and require a large number of training tasks. To address these challenges, we propose a novel approach for learning a policy committee that includes at least one near-optimal policy with high probability for tasks encountered during execution. While we show that this problem is in general inapproximable, we present two practical algorithmic solutions. The first yields provable approximation and task sample complexity guarantees when tasks are low-dimensional (the best we can do due to inapproximability), whereas the second is a general and practical gradient-based approach. In addition, we provide a provable sample complexity bound for few-shot learning. Our experiments on MuJoCo and Meta-World show that the proposed approach outperforms state-of-the-art multi-task, meta-, and task clustering baselines in training, generalization, and few-shot learning, often by a large margin. Our code is available at https://github.com/CERL-WUSTL/PACMAN. | [
{
"version": "v1",
"created": "Wed, 26 Feb 2025 22:45:25 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 21:04:56 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ge",
"Luise",
""
],
[
"Lanier",
"Michael",
""
],
[
"Sarkar",
"Anindya",
""
],
[
"Guresti",
"Bengisu",
""
],
[
"Zhang",
"Chongjie",
""
],
[
"Vorobeychik",
"Yevgeniy",
""
]
] |
2503.02489 | Victor Gaydamachenko | Victor Gaydamachenko, Christoph Kissling, Lukas Gr\"unhaupt | An rf-SQUID-based traveling-wave parametric amplifier with -84 dBm input saturation power across more than one octave bandwidth | null | null | 10.1103/1qk4-fzkq | null | quant-ph cond-mat.supr-con | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traveling-wave parametric amplifiers (TWPAs) have become an essential tool for the readout of quantum circuits and the search for dark matter. We report on the implementation of an rf-SQUID-based Josephson TWPA with an average saturation power of -84 dBm, while providing an average power gain of 20 dB from 3.5 to 8.5 GHz. This wide bandwidth is enabled by reducing the curvature of the dispersion in the signal band while suppressing detrimental mixing processes. With 2393 rf-SQUIDs our device has a comparable number of nonlinear elements, but achieves ten times higher saturation power than previous works, showing that using rf-SQUIDs results in a more favorable trade-off between device length and saturation power. Harnessing wideband characterization techniques we determine the TWPA's excess noise above the quantum limit to be 0.8 to 1.5 photons. In addition, we validate signal-to-noise ratio improvement as a practical measure for tuning the TWPA to minimize the total system noise, and demonstrate that its optimum does not coincide with the highest gain. | [
{
"version": "v1",
"created": "Tue, 04 Mar 2025 10:55:10 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Gaydamachenko",
"Victor",
""
],
[
"Kissling",
"Christoph",
""
],
[
"Grünhaupt",
"Lukas",
""
]
] |
2503.02857 | Nuria Chandra | Nuria Alina Chandra, Ryan Murtfeldt, Lin Qiu, Arnab Karmakar, Hannah Lee, Emmanuel Tanumihardja, Kevin Farhat, Ben Caffee, Sejin Paik, Changyeon Lee, Jongwook Choi, Aerin Kim, Oren Etzioni | Deepfake-Eval-2024: A Multi-Modal In-the-Wild Benchmark of Deepfakes Circulated in 2024 | null | null | null | null | cs.CV cs.AI cs.CY | http://creativecommons.org/licenses/by-sa/4.0/ | In the age of increasingly realistic generative AI, robust deepfake detection is essential for mitigating fraud and disinformation. While many deepfake detectors report high accuracy on academic datasets, we show that these academic benchmarks are out of date and not representative of real-world deepfakes. We introduce Deepfake-Eval-2024, a new deepfake detection benchmark consisting of in-the-wild deepfakes collected from social media and deepfake detection platform users in 2024. Deepfake-Eval-2024 consists of 45 hours of videos, 56.5 hours of audio, and 1,975 images, encompassing the latest manipulation technologies. The benchmark contains diverse media content from 88 different websites in 52 different languages. We find that the performance of open-source state-of-the-art deepfake detection models drops precipitously when evaluated on Deepfake-Eval-2024, with AUC decreasing by 50% for video, 48% for audio, and 45% for image models compared to previous benchmarks. We also evaluate commercial deepfake detection models and models finetuned on Deepfake-Eval-2024, and find that they have superior performance to off-the-shelf open-source models, but do not yet reach the accuracy of deepfake forensic analysts. The dataset is available at https://github.com/nuriachandra/Deepfake-Eval-2024. | [
{
"version": "v1",
"created": "Tue, 04 Mar 2025 18:33:22 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Mar 2025 20:24:16 GMT"
},
{
"version": "v3",
"created": "Mon, 24 Mar 2025 20:46:15 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 16:27:56 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chandra",
"Nuria Alina",
""
],
[
"Murtfeldt",
"Ryan",
""
],
[
"Qiu",
"Lin",
""
],
[
"Karmakar",
"Arnab",
""
],
[
"Lee",
"Hannah",
""
],
[
"Tanumihardja",
"Emmanuel",
""
],
[
"Farhat",
"Kevin",
""
],
[
"Caffee",
"Ben",
""
],
[
"Paik",
"Sejin",
""
],
[
"Lee",
"Changyeon",
""
],
[
"Choi",
"Jongwook",
""
],
[
"Kim",
"Aerin",
""
],
[
"Etzioni",
"Oren",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.