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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2410.14363 | Anushka De Ms | Tathagata Banerjee, Anushka De, Subhamoy Maitra, Diganta Mukherjee | Skill vs. Chance Quantification for Popular Card & Board Games | 25 pages, 9 figures | null | null | null | cs.GT stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a data-driven statistical framework to quantify the role of skill in games, addressing the long-standing question of whether success in a game is predominantly driven by skill or chance. We analyze player level data from four popular games Chess, Rummy, Ludo, and Teen Patti, using empirical win statistics across varying levels of experience. By modeling win rate as a function of experience through a regression framework and employing empirical bootstrap resampling, we estimate the degree to which outcomes improve with repeated play. To summarize these dynamics, we propose a flexible skill score that emphasizes learning over initial performance, aligning with practical and regulatory interpretations of skill. Our results reveal a clear ranking, with Chess showing the highest skill component and Teen Patti the lowest, while Rummy and Ludo fall in between. The proposed framework is transparent, reproducible, and adaptable to other game formats and outcome metrics, offering potential applications in legal classification, game design, and player performance analysis. | [
{
"version": "v1",
"created": "Fri, 18 Oct 2024 10:49:54 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 06:43:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Banerjee",
"Tathagata",
""
],
[
"De",
"Anushka",
""
],
[
"Maitra",
"Subhamoy",
""
],
[
"Mukherjee",
"Diganta",
""
]
] |
2410.14623 | Ce Sui | Ce Sui, Deaglan J. Bartlett, Shivam Pandey, Harry Desmond, Pedro G. Ferreira, Benjamin D. Wandelt | syren-new: Precise formulae for the linear and nonlinear matter power spectra with massive neutrinos and dynamical dark energy | 18 pages, 15 figures | A&A 698, A1 (2025) | 10.1051/0004-6361/202452854 | null | astro-ph.CO astro-ph.IM cs.LG cs.NE | http://creativecommons.org/licenses/by/4.0/ | Current and future large scale structure surveys aim to constrain the neutrino mass and the equation of state of dark energy. We aim to construct accurate and interpretable symbolic approximations to the linear and nonlinear matter power spectra as a function of cosmological parameters in extended $\Lambda$CDM models which contain massive neutrinos and non-constant equations of state for dark energy. This constitutes an extension of the syren-halofit emulators to incorporate these two effects, which we call syren-new (SYmbolic-Regression-ENhanced power spectrum emulator with NEutrinos and $W_0-w_a$). We also obtain a simple approximation to the derived parameter $\sigma_8$ as a function of the cosmological parameters for these models. Our results for the linear power spectrum are designed to emulate CLASS, whereas for the nonlinear case we aim to match the results of EuclidEmulator2. We compare our results to existing emulators and $N$-body simulations. Our analytic emulators for $\sigma_8$, the linear and nonlinear power spectra achieve root mean squared errors of 0.1%, 0.3% and 1.3%, respectively, across a wide range of cosmological parameters, redshifts and wavenumbers. We verify that emulator-related discrepancies are subdominant compared to observational errors and other modelling uncertainties when computing shear power spectra for LSST-like surveys. Our expressions have similar accuracy to existing (numerical) emulators, but are at least an order of magnitude faster, both on a CPU and GPU. Our work greatly improves the accuracy, speed and range of applicability of current symbolic approximations to the linear and nonlinear matter power spectra. We provide publicly available code for all symbolic approximations found. | [
{
"version": "v1",
"created": "Fri, 18 Oct 2024 17:22:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Sui",
"Ce",
""
],
[
"Bartlett",
"Deaglan J.",
""
],
[
"Pandey",
"Shivam",
""
],
[
"Desmond",
"Harry",
""
],
[
"Ferreira",
"Pedro G.",
""
],
[
"Wandelt",
"Benjamin D.",
""
]
] |
2410.15332 | Junhao Hu | Junhao Hu, Wenrui Huang, Weidong Wang, Haoyi Wang, Tiancheng Hu, Qin Zhang, Hao Feng, Xusheng Chen, Yizhou Shan, Tao Xie | EPIC: Efficient Position-Independent Caching for Serving Large Language Models | null | null | null | null | cs.LG cs.CL cs.DC cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) show great capabilities in a wide range of applications, but serving them efficiently becomes increasingly challenging as requests (prompts) become more complex. Context caching improves serving performance by reusing Key-Value (KV) vectors, the intermediate representations of tokens that are repeated across requests. However, existing context caching requires exact prefix matches across requests, limiting reuse cases in settings such as few-shot learning and retrieval-augmented generation, where immutable content (e.g., documents) remains unchanged across requests but is preceded by varying prefixes. Position-Independent Caching (PIC) addresses this issue by enabling modular reuse of the KV vectors regardless of prefixes. We formalize PIC and advance prior work by introducing EPIC, a serving system incorporating our new LegoLink algorithm, which mitigates the inappropriate "attention sink" effect at every document beginning, to maintain accuracy with minimal computation. Experiments show that EPIC achieves up to 8x improvements in Time-To-First-Token (TTFT) and 7x throughput gains over existing systems, with negligible or no accuracy loss. | [
{
"version": "v1",
"created": "Sun, 20 Oct 2024 08:42:29 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Feb 2025 14:50:00 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 09:24:50 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Hu",
"Junhao",
""
],
[
"Huang",
"Wenrui",
""
],
[
"Wang",
"Weidong",
""
],
[
"Wang",
"Haoyi",
""
],
[
"Hu",
"Tiancheng",
""
],
[
"Zhang",
"Qin",
""
],
[
"Feng",
"Hao",
""
],
[
"Chen",
"Xusheng",
""
],
[
"Shan",
"Yizhou",
""
],
[
"Xie",
"Tao",
""
]
] |
2410.16354 | Natalie Allen | Natalie Allen, Pascal A. Oesch, Sune Toft, Jasleen Matharu, Conor J. R. McPartland, Andrea Weibel, Gabe Brammer, Rebecca A. A. Bowler, Kei Ito, Rashmi Gottumukkala, Francesca Rizzo, Francesco Valentino, Rohan G. Varadaraj, John R. Weaver, Katherine E. Whitaker | Galaxy Size and Mass Build-up in the First 2 Gyrs of Cosmic History from Multi-Wavelength JWST NIRCam Imaging | 14 pages, 9 figures, submitted to A&A | A&A 698, A30 (2025) | 10.1051/0004-6361/202452690 | null | astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The evolution of galaxy sizes in different wavelengths provides unique insights on galaxy build-up across cosmic epochs. Such measurements can now finally be done at $z>3$ thanks to the exquisite spatial resolution and multi-wavelength capability of the JWST. With the public data from the CEERS, PRIMER-UDS, and PRIMER-COSMOS surveys, we measure the sizes of $\sim 3500$ star-forming galaxies at $3 \leqslant z<9$, in 7 NIRCam bands using the multi-wavelength model fitting code GalfitM. The size-mass relation is measured in four redshift bins, across all NIRCam bands. We find that, the slope and intrinsic scatter of the rest-optical size-mass relation are constant across this redshift range and consistent with previous HST-based studies at low-z. When comparing the relations across different wavelengths, the average rest-optical and rest-UV relations are consistent with each other up to $z=6$, but the intrinsic scatter is largest in rest-UV wavelengths compared to rest-optical and redder bands. This behaviour is independent of redshift and we speculate that it is driven by bursty star-formation in $z>4$ galaxies. Additionally, for $3\leqslant z<4$ star-forming galaxies at $\rm M_* > 10^{10} M_{\odot}$, we find smaller rest-$\rm 1\rm\,\mu m$ sizes in comparison to rest-optical (and rest-UV) sizes, suggestive of colour gradients. When comparing to simulations, we find agreement over $\rm M_* \approx 10^{9} - 10^{10} M_{\odot}$ but beyond this mass, the observed size-mass relation is significantly steeper. Our results show the power of JWST/NIRCam to provide new constraints on galaxy formation models. | [
{
"version": "v1",
"created": "Mon, 21 Oct 2024 18:00:00 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Allen",
"Natalie",
""
],
[
"Oesch",
"Pascal A.",
""
],
[
"Toft",
"Sune",
""
],
[
"Matharu",
"Jasleen",
""
],
[
"McPartland",
"Conor J. R.",
""
],
[
"Weibel",
"Andrea",
""
],
[
"Brammer",
"Gabe",
""
],
[
"Bowler",
"Rebecca A. A.",
""
],
[
"Ito",
"Kei",
""
],
[
"Gottumukkala",
"Rashmi",
""
],
[
"Rizzo",
"Francesca",
""
],
[
"Valentino",
"Francesco",
""
],
[
"Varadaraj",
"Rohan G.",
""
],
[
"Weaver",
"John R.",
""
],
[
"Whitaker",
"Katherine E.",
""
]
] |
2410.17200 | Guodong Pang | Guodong Pang and Etienne Pardoux | SPDE for stochastic SIR epidemic models with infection-age dependent infectivity | null | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the stochastic SIR epidemic model with infection-age dependent infectivity for which a measure-valued process is used to describe the ages of infection for each individual. We establish a functional law of large numbers (FLLN) and a functional central limit theorem (FCLT) for the properly scaled measure-valued processes together with the other epidemic processes to describe the evolution dynamics. In the FLLN, assuming that the hazard rate function of the infection periods is bounded and the ages at time 0 of the infections of the initially infected individuals are bounded, we obtain a PDE limit for the LLN-scaled measure-valued process, for which we characterize its solution explicitly. The PDE is linear with a boundary condition given by the unique solution to a set of Volterra-type nonlinear integral equations. In the FCLT, we obtain an SPDE for the CLT-scaled measure-valued process, driven by two independent white noises coming from the infection and recovery processes. The SPDE is also linear and coupled with the solution to a system of stochastic Volterra-type linear integral equations driven by three independent Gaussian noises, one from the random infection functions in addition to the two white noises mentioned above. The solution to the SPDE can be also explicitly characterized, given this auxiliary process. The uniqueness of the SPDE solution is established under stronger assumptions (density and its derivative being locally bounded) on the distribution function of an infectious duration. | [
{
"version": "v1",
"created": "Tue, 22 Oct 2024 17:19:27 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 09:54:50 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Pang",
"Guodong",
""
],
[
"Pardoux",
"Etienne",
""
]
] |
2410.17547 | Sharath Matada | Sharath Matada, Luke Bhan, Yuanyuan Shi, Nikolay Atanasov | Generalizable Motion Planning via Operator Learning | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | In this work, we introduce a planning neural operator (PNO) for predicting the value function of a motion planning problem. We recast value function approximation as learning a single operator from the cost function space to the value function space, which is defined by an Eikonal partial differential equation (PDE). Therefore, our PNO model, despite being trained with a finite number of samples at coarse resolution, inherits the zero-shot super-resolution property of neural operators. We demonstrate accurate value function approximation at $16\times$ the training resolution on the MovingAI lab's 2D city dataset, compare with state-of-the-art neural value function predictors on 3D scenes from the iGibson building dataset and showcase optimal planning with 4-DOF robotic manipulators. Lastly, we investigate employing the value function output of PNO as a heuristic function to accelerate motion planning. We show theoretically that the PNO heuristic is $\epsilon$-consistent by introducing an inductive bias layer that guarantees our value functions satisfy the triangle inequality. With our heuristic, we achieve a $30\%$ decrease in nodes visited while obtaining near optimal path lengths on the MovingAI lab 2D city dataset, compared to classical planning methods ($A^\ast$, $RRT^\ast$). | [
{
"version": "v1",
"created": "Wed, 23 Oct 2024 04:06:35 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Mar 2025 05:11:18 GMT"
},
{
"version": "v3",
"created": "Mon, 26 May 2025 21:03:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Matada",
"Sharath",
""
],
[
"Bhan",
"Luke",
""
],
[
"Shi",
"Yuanyuan",
""
],
[
"Atanasov",
"Nikolay",
""
]
] |
2410.17573 | Xi Li | Xiaohuan Bi, Xi Li | Securing Federated Learning against Backdoor Threats with Foundation Model Integration | This paper has been accepted to IEEE IRI 2025 | null | null | null | cs.LG cs.CR | http://creativecommons.org/licenses/by/4.0/ | Federated Learning (FL) enables decentralized model training while preserving privacy. Recently, the integration of Foundation Models (FMs) into FL has enhanced performance but introduced a novel backdoor attack mechanism. Attackers can exploit FM vulnerabilities to embed backdoors into synthetic data generated by FMs. During global model fusion, these backdoors are transferred to the global model through compromised synthetic data, subsequently infecting all client models. Existing FL backdoor defenses are ineffective against this novel attack due to its fundamentally different mechanism compared to classic ones. In this work, we propose a novel data-free defense strategy that addresses both classic and novel backdoor attacks in FL. The shared attack pattern lies in the abnormal activations within the hidden feature space during model aggregation. Hence, we propose to constrain internal activations to remain within reasonable ranges, effectively mitigating attacks while preserving model functionality. The activation constraints are optimized using synthetic data alongside FL training. Extensive experiments demonstrate its effectiveness against both novel and classic backdoor attacks, outperforming existing defenses. | [
{
"version": "v1",
"created": "Wed, 23 Oct 2024 05:54:41 GMT"
},
{
"version": "v2",
"created": "Fri, 16 May 2025 19:15:40 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 16:10:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bi",
"Xiaohuan",
""
],
[
"Li",
"Xi",
""
]
] |
2410.17844 | Nat\'alia Rodrigues | Nat\'alia V. N. Rodrigues, Natal\'i S. M. de Santi, L. Raul Abramo, Antonio D. Montero-Dorta | Exploring the halo-galaxy connection with probabilistic approaches | 14 pages, 8 figures, 6 tables | A&A 698, A3 (2025) | 10.1051/0004-6361/202453284 | null | astro-ph.CO astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The connection between galaxies and dark matter halos encompasses a range of processes and play a pivotal role in our understanding of galaxy formation and evolution. Traditionally, this link has been established through physical or empirical models. Machine learning techniques are adaptable tools that handle high-dimensional data and grasp associations between numerous attributes. In particular, probabilistic models capture the stochasticity inherent to these complex relations. We compare different probabilistic machine learning methods to model the uncertainty in the halo-galaxy connection and efficiently generate galaxy catalogs that faithfully resemble the reference sample by predicting joint distributions of central galaxy properties conditioned to their host halo features. The analysis is based on the IllustrisTNG300 simulation. The methods model the distributions in different ways. We compare a multilayer perceptron that predicts the parameters of a multivariate Gaussian distribution, a multilayer perceptron classifier, and the method of normalizing flows. The classifier predicts the parameters of a Categorical distribution, which are defined in a high-dimensional parameter space through a Voronoi cell-based hierarchical scheme. We evaluate the model's performances under various sample selections based on halo properties. The three methods exhibit comparable results, with normalizing flows showing the best performance in most scenarios. The models reproduce the main features of galaxy properties distributions with high-fidelity and reproduce the results obtained with traditional, deterministic, estimators. Our results also indicate that different halos and galaxy populations are subject to varying degrees of stochasticity, which has relevant implications for studies of large-scale structure. | [
{
"version": "v1",
"created": "Wed, 23 Oct 2024 13:09:43 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Rodrigues",
"Natália V. N.",
""
],
[
"de Santi",
"Natalí S. M.",
""
],
[
"Abramo",
"L. Raul",
""
],
[
"Montero-Dorta",
"Antonio D.",
""
]
] |
2410.18122 | Ivo Verhoeven | Ivo Verhoeven, Pushkar Mishra, Ekaterina Shutova | Yesterday's News: Benchmarking Multi-Dimensional Out-of-Distribution Generalization of Misinformation Detection Models | Under review | null | null | null | cs.IR cs.CL | http://creativecommons.org/licenses/by/4.0/ | This article introduces misinfo-general, a benchmark dataset for evaluating misinformation models' ability to perform out-of-distribution generalization. Misinformation changes rapidly, much more quickly than moderators can annotate at scale, resulting in a shift between the training and inference data distributions. As a result, misinformation detectors need to be able to perform out-of-distribution generalization, an attribute they currently lack. Our benchmark uses distant labelling to enable simulating covariate shifts in misinformation content. We identify time, event, topic, publisher, political bias, misinformation type as important axes for generalization, and we evaluate a common class of baseline models on each. Using article metadata, we show how this model fails desiderata, which is not necessarily obvious from classification metrics. Finally, we analyze properties of the data to ensure limited presence of modelling shortcuts. We make the dataset and accompanying code publicly available: https://github.com/ioverho/misinfo-general | [
{
"version": "v1",
"created": "Sat, 12 Oct 2024 09:46:36 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 19:18:59 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Verhoeven",
"Ivo",
""
],
[
"Mishra",
"Pushkar",
""
],
[
"Shutova",
"Ekaterina",
""
]
] |
2410.18155 | Ricardo Gallego Torrom\'e | Ricardo Gallego Torrom\'e | On the Unruh effect for hyperbolic observers in spacetimes of maximal proper acceleration | 22 pages | null | null | null | gr-qc hep-th math-ph math.MP physics.acc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, the notion of spacetime of maximal proper acceleration is motivated as a weak form to implement general covariance and a generalized form of Einstein's equivalence principle from a physical point of view and the fundamental geometric and kinematic properties of such spaces are discussed. Thereafter the Unruh temperature formula is generalized to the case of hyperbolic observers in spacetimes of maximal proper acceleration. Such a generalization implies the existence of a maximal value for the Unruh temperature. We discuss this result for an electrodynamic model of point charged particles in a spacetime of maximal proper acceleration. It is shown that according to the model, the maximal Unruh temperature must be of order $4.8\cdot 10^{10}/N$ K for current high acceleration electron laser-plasma acceleration systems, where $N$ is the population of the typical bunch. | [
{
"version": "v1",
"created": "Wed, 23 Oct 2024 08:15:03 GMT"
},
{
"version": "v2",
"created": "Tue, 05 Nov 2024 13:02:07 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Dec 2024 16:37:52 GMT"
},
{
"version": "v4",
"created": "Mon, 16 Dec 2024 14:25:14 GMT"
},
{
"version": "v5",
"created": "Tue, 04 Feb 2025 14:10:52 GMT"
},
{
"version": "v6",
"created": "Tue, 27 May 2025 14:21:49 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Torromé",
"Ricardo Gallego",
""
]
] |
2410.18693 | Yuyang Ding | Yuyang Ding, Xinyu Shi, Xiaobo Liang, Juntao Li, Zhaopeng Tu, Qiaoming Zhu, Min Zhang | Unleashing LLM Reasoning Capability via Scalable Question Synthesis from Scratch | ACL 2025 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Improving the mathematical reasoning capabilities of Large Language Models (LLMs) is critical for advancing artificial intelligence. However, access to extensive, diverse, and high-quality reasoning datasets remains a significant challenge, particularly for the open-source community. In this paper, we propose ScaleQuest, a novel, scalable, and cost-effective data synthesis method that enables the generation of large-scale mathematical reasoning datasets using lightweight 7B-scale models. ScaleQuest introduces a two-stage question-tuning process comprising Question Fine-Tuning (QFT) and Question Preference Optimization (QPO) to unlock the question generation capabilities of problem-solving models. By generating diverse questions from scratch -- without relying on powerful proprietary models or seed data -- we produce a dataset of 1 million problem-solution pairs. Our experiments demonstrate that models trained on our data outperform existing open-source datasets in both in-domain and out-of-domain evaluations. Furthermore, our approach shows continued performance improvement as the volume of training data increases, highlighting its potential for ongoing data scaling. The extensive improvements observed in code reasoning tasks demonstrate the generalization capabilities of our proposed method. Our work provides the open-source community with a practical solution to enhance the mathematical reasoning abilities of LLMs. | [
{
"version": "v1",
"created": "Thu, 24 Oct 2024 12:42:04 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 11:56:56 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ding",
"Yuyang",
""
],
[
"Shi",
"Xinyu",
""
],
[
"Liang",
"Xiaobo",
""
],
[
"Li",
"Juntao",
""
],
[
"Tu",
"Zhaopeng",
""
],
[
"Zhu",
"Qiaoming",
""
],
[
"Zhang",
"Min",
""
]
] |
2410.19049 | Riccardo Comi | Sergio Benvenuti, Riccardo Comi, Sara Pasquetti | Star-triangle dualities and supersymmetric improved bifundamentals | 69+15 pages with many figures, added references in v2 | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | Recently it was shown that mirror duals of 3d and 4d theories with four super-charges can be described by generalized quiver theories, constructed using strongly coupled SCFTs as elementary building blocks that replace and improve standard bifundamentals. In this work we study and extend the family of such improved bifundamentals and discuss the network of star-triangle dualities they satisfy. We provide a field theoretic proof of the star-triangle dualities, which only assumes the basic Seiberg dualities, using the sequential deconfinement technique. | [
{
"version": "v1",
"created": "Thu, 24 Oct 2024 18:00:06 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 09:43:22 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Benvenuti",
"Sergio",
""
],
[
"Comi",
"Riccardo",
""
],
[
"Pasquetti",
"Sara",
""
]
] |
2410.21013 | Akhilesh Kakolu Ramarao | Akhilesh Kakolu Ramarao, Kevin Tang, Dinah Baer-Henney | Frequency matters: Modeling irregular morphological patterns in Spanish with Transformers | Typos and grammatical corrections | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Over the past decade, various studies have addressed how speakers solve the so-called `The Paradigm Cell Filling Problem' (PCFP) \citep{ackerman2009parts} across different languages. The PCFP addresses a fundamental question in morphological processing: how do speakers accurately generate inflected forms of words when presented with incomplete paradigms? This problem is particularly salient when modeling complex inflectional systems. We focus on Spanish verbal paradigms, where certain verbs follow an irregular L-shaped pattern, where the first-person singular present indicative stem matches the stem used throughout the present subjunctive mood. We formulate the problem as a morphological reinflection task. Specifically, we investigate the role of input frequency in the acquisition of regular versus irregular L-shaped patterns in transformer models. By systematically manipulating the input distributions and analyzing model behavior, we reveal four key findings: 1) Models perform better on L-shaped verbs compared to regular verbs, especially in uneven frequency conditions; 2) Robust primacy effects are observed, but no consistent recency effects; 3) Memorization becomes more prominent as the proportion of L-shaped verbs increases; 4) There is a tendency to regularize L-shaped verbs when their consonant alternation pairs are rare or absent in the training data. | [
{
"version": "v1",
"created": "Mon, 28 Oct 2024 13:36:46 GMT"
},
{
"version": "v2",
"created": "Fri, 13 Dec 2024 12:27:37 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Feb 2025 10:15:07 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 15:48:35 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ramarao",
"Akhilesh Kakolu",
""
],
[
"Tang",
"Kevin",
""
],
[
"Baer-Henney",
"Dinah",
""
]
] |
2410.21296 | Serge Dolgikh | Serge Dolgikh | The Trap of Presumed Equivalence: Artificial General Intelligence Should Not Be Assessed on the Scale of Human Intelligence | null | Discov Artif Intell 5, 70 (2025) | 10.1007/s44163-025-00317-6 | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A traditional approach to assessing emerging intelligence in the theory of intelligent systems is based on the similarity, "imitation" of human-like actions and behaviors, benchmarking the performance of intelligent systems on the scale of human cognitive skills. In this work we attempt to outline the shortcomings of this line of thought, which is based on the implicit presumption of the equivalence and compatibility of the originating and emergent intelligences. We provide arguments to the point that under some natural assumptions, developing intelligent systems will be able to form their own intents and objectives. Then, the difference in the rate of progress of natural and artificial systems that was noted on multiple occasions in the discourse on artificial intelligence can lead to the scenario of a progressive divergence of the intelligences, in their cognitive abilities, functions and resources, values, ethical frameworks, worldviews, intents and existential objectives: the scenario of the AGI evolutionary gap. We discuss evolutionary processes that can guide the development of emergent intelligent systems and attempt to identify the starting point of the progressive divergence scenario. | [
{
"version": "v1",
"created": "Mon, 14 Oct 2024 13:39:58 GMT"
},
{
"version": "v2",
"created": "Mon, 11 Nov 2024 13:43:59 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Dolgikh",
"Serge",
""
]
] |
2410.21952 | Emanuele Ledda | Emanuele Ledda, Giovanni Scodeller, Daniele Angioni, Giorgio Piras, Antonio Emanuele Cin\`a, Giorgio Fumera, Battista Biggio and Fabio Roli | On the Robustness of Adversarial Training Against Uncertainty Attacks | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In learning problems, the noise inherent to the task at hand hinders the possibility to infer without a certain degree of uncertainty. Quantifying this uncertainty, regardless of its wide use, assumes high relevance for security-sensitive applications. Within these scenarios, it becomes fundamental to guarantee good (i.e., trustworthy) uncertainty measures, which downstream modules can securely employ to drive the final decision-making process. However, an attacker may be interested in forcing the system to produce either (i) highly uncertain outputs jeopardizing the system's availability or (ii) low uncertainty estimates, making the system accept uncertain samples that would instead require a careful inspection (e.g., human intervention). Therefore, it becomes fundamental to understand how to obtain robust uncertainty estimates against these kinds of attacks. In this work, we reveal both empirically and theoretically that defending against adversarial examples, i.e., carefully perturbed samples that cause misclassification, additionally guarantees a more secure, trustworthy uncertainty estimate under common attack scenarios without the need for an ad-hoc defense strategy. To support our claims, we evaluate multiple adversarial-robust models from the publicly available benchmark RobustBench on the CIFAR-10 and ImageNet datasets. | [
{
"version": "v1",
"created": "Tue, 29 Oct 2024 11:12:44 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:41:42 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ledda",
"Emanuele",
""
],
[
"Scodeller",
"Giovanni",
""
],
[
"Angioni",
"Daniele",
""
],
[
"Piras",
"Giorgio",
""
],
[
"Cinà",
"Antonio Emanuele",
""
],
[
"Fumera",
"Giorgio",
""
],
[
"Biggio",
"Battista",
""
],
[
"Roli",
"Fabio",
""
]
] |
2410.22275 | Janne Junnila | Mario Bonk, Janne Junnila, Steffen Rohde, Yilin Wang | Piecewise geodesic Jordan curves II: Loewner energy, projective structures, and accessory parameters | 38 pages, 7 figures; v2 with updated introduction and some minor changes | null | null | null | math.CV math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we consider Jordan curves on the Riemann sphere passing through $n \ge 3$ given points. We show that in each relative isotopy class of such curves, there exists a unique curve that minimizes the Loewner energy. These curves have the property that each arc between two consecutive points is a hyperbolic geodesic in the domain bounded by the other arcs. This geodesic property lets us define a complex projective structure whose holonomy lies in $\mathrm{PSL}(2,\mathbb{R})$. We show that the quadratic differential comparing this projective structure to the trivial projective structure on the sphere has simple poles whose residues (accessory parameters) are given by the Wirtinger derivatives of the minimal Loewner energy. This is reminiscent of Polyakov's conjecture for Fuchsian projective structures, proven by Takhtajan and Zograf. Finally, we show that the projective structures we obtain are related to Fuchsian projective structures through $\pi$-grafting. | [
{
"version": "v1",
"created": "Tue, 29 Oct 2024 17:30:05 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 14:05:05 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bonk",
"Mario",
""
],
[
"Junnila",
"Janne",
""
],
[
"Rohde",
"Steffen",
""
],
[
"Wang",
"Yilin",
""
]
] |
2410.22768 | Takuto Yamamoto | Takuto Yamamoto, Hirosato Akahoshi, Shigeru Kitazawa | Emergence of Human-Like Attention in Self-Supervised Vision Transformers: an eye-tracking study | null | Neural Networks, 107595, 2025 | 10.1016/j.neunet.2025.107595 | null | q-bio.NC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Many models of visual attention have been proposed so far. Traditional bottom-up models, like saliency models, fail to replicate human gaze patterns, and deep gaze prediction models lack biological plausibility due to their reliance on supervised learning. Vision Transformers (ViTs), with their self-attention mechanisms, offer a new approach but often produce dispersed attention patterns if trained with supervised learning. This study explores whether self-supervised DINO (self-DIstillation with NO labels) training enables ViTs to develop attention mechanisms resembling human visual attention. Using video stimuli to capture human gaze dynamics, we found that DINO-trained ViTs closely mimic human attention patterns, while those trained with supervised learning deviate significantly. An analysis of self-attention heads revealed three distinct clusters: one focusing on foreground objects, one on entire objects, and one on the background. DINO-trained ViTs offer insight into how human overt attention and figure-ground separation develop in visual perception. | [
{
"version": "v1",
"created": "Wed, 30 Oct 2024 07:37:19 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Yamamoto",
"Takuto",
""
],
[
"Akahoshi",
"Hirosato",
""
],
[
"Kitazawa",
"Shigeru",
""
]
] |
2410.22881 | Imad Ali Shah | Imad Ali Shah, Fahad Mumtaz Malik, Muhammad Waqas Ashraf | SFA-UNet: More Attention to Multi-Scale Contrast and Contextual Information in Infrared Small Object Segmentation | Accepted and Presented at PRIP 2023 | In 2023 Pattern Recognition and Information Processing (PRIP), pp.147-152 | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Computer vision researchers have extensively worked on fundamental infrared visual recognition for the past few decades. Among various approaches, deep learning has emerged as the most promising candidate. However, Infrared Small Object Segmentation (ISOS) remains a major focus due to several challenges including: 1) the lack of effective utilization of local contrast and global contextual information; 2) the potential loss of small objects in deep models; and 3) the struggling to capture fine-grained details and ignore noise. To address these challenges, we propose a modified U-Net architecture, named SFA-UNet, by combining Scharr Convolution (SC) and Fast Fourier Convolution (FFC) in addition to vertical and horizontal Attention gates (AG) into UNet. SFA-UNet utilizes double convolution layers with the addition of SC and FFC in its encoder and decoder layers. SC helps to learn the foreground-to-background contrast information whereas FFC provide multi-scale contextual information while mitigating the small objects vanishing problem. Additionally, the introduction of vertical AGs in encoder layers enhances the model's focus on the targeted object by ignoring irrelevant regions. We evaluated the proposed approach on publicly available, SIRST and IRSTD datasets, and achieved superior performance by an average 0.75% with variance of 0.025 of all combined metrics in multiple runs as compared to the existing state-of-the-art methods | [
{
"version": "v1",
"created": "Wed, 30 Oct 2024 10:21:23 GMT"
},
{
"version": "v2",
"created": "Sat, 16 Nov 2024 14:10:33 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Shah",
"Imad Ali",
""
],
[
"Malik",
"Fahad Mumtaz",
""
],
[
"Ashraf",
"Muhammad Waqas",
""
]
] |
2410.23302 | Hessam Bakhshi-Khaniki | Hessam Bakhshi-Khaniki, Reza Tavakkoli-Moghaddam, Zdenek Hanzalek, Behdin Vahedi-Nouri | A Flexible Job Shop Scheduling Problem Involving Reconfigurable Machine Tools Under Industry 5.0 | 14 pages | Advances in Production Management Systems. Production Management Systems for Volatile, Uncertain, Complex, and Ambiguous Environments. APMS 2024. IFIP Advances in Information and Communication Technology, vol 733. Springer, Cham | 10.1007/978-3-031-71645-4_17 | null | cs.NE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The rise of Industry 5.0 has introduced new demands for manufacturing companies, requiring a shift in how production schedules are managed to address human centered, environmental, and economic goals comprehensively. The flexible job shop scheduling problem (FJSSP), which involves processing operations on various capable machines, accurately reflects the complexities of modern manufacturing settings. This paper investigates the FJSSP involving reconfigurable machine tools with configuration dependent setup times, while integrating human aspects like worker assignments, moving time, and rest periods, as well as minimizing total energy consumption. A mixed-integer programming (MIP) model is developed to simultaneously optimize these objectives. The model determines the assignment of operations to machines, workers, and configurations while sequencing operations, scheduling worker movements, and respecting rest periods, and minimizing overall energy consumption. Given the NPhard nature of the FJSSP with worker assignments and reconfigurable tools, a memetic algorithm (MA) is proposed. This metaheuristic evolutionary algorithm features a three layer chromosome encoding method, specialized crossover and mutation strategies, and neighborhood search mechanisms to enhance solution quality and diversity. Comparisons of MA with MIP and genetic algorithms (GA) on benchmark instances demonstrate the MA efficiency and effectiveness, particularly for larger problem instances where MIP becomes impractical. This research paves the way for sustainable and resilient production schedules tailored for the factory of the future under the Industry 5.0 paradigm. The work bridges a crucial gap in current literature by integrating worker and environmental impact into the FJSSP with reconfigurable machine models. | [
{
"version": "v1",
"created": "Wed, 16 Oct 2024 11:40:06 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 12:27:02 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bakhshi-Khaniki",
"Hessam",
""
],
[
"Tavakkoli-Moghaddam",
"Reza",
""
],
[
"Hanzalek",
"Zdenek",
""
],
[
"Vahedi-Nouri",
"Behdin",
""
]
] |
2410.23353 | Yoshifumi Nakata Dr | Yoshifumi Nakata, Yuki Takeuchi, Martin Kliesch, Andrew Darmawan | On computational complexity of unitary and state design properties | Ver 1. 22 pages, 1 figure, and 1 table. Ver 2. 27 pages, 1 figure, and 1 table. Results are updated and implications are added. Ver 3. 33 pages, 2 figures, and 1 table. An error in a previous proof is fixed | null | null | YITP-24-141 | quant-ph cond-mat.stat-mech hep-th math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We investigate unitary and state $t$-designs from a computational complexity perspective. First, we address the problems of computing frame potentials that characterize (approximate) $t$-designs. We present a quantum algorithm for computing frame potentials and establish the following: (1) exact computation can be achieved by a single query to a $\# \textsf{P}$-oracle and is $\# \textsf{P}$-hard; (2) for state vectors, deciding whether the frame potential is larger than or smaller than certain values is $\textsf{BQP}$-complete, provided the promise gap between the two values is inverse-polynomial in the number of qubits; and (3) for both state vectors and unitaries, this promise problem is $\textsf{PP}$-complete if the promise gap is exponentially small. Second, we address the promise problem of deciding whether or not a given set is a good approximation to a design. Given a certain promise gap that could be constant, we show that this problem is $\textsf{PP}$-hard, highlighting the inherent computational difficulty of determining properties of unitary and state designs. We further identify the implications of our results across diverse areas, including variational methods for constructing designs, diagnosing quantum chaos through out-of-time-ordered correlators (OTOCs), and exploring emergent designs in Hamiltonian systems. | [
{
"version": "v1",
"created": "Wed, 30 Oct 2024 18:00:35 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Jan 2025 00:52:16 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 04:53:39 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Nakata",
"Yoshifumi",
""
],
[
"Takeuchi",
"Yuki",
""
],
[
"Kliesch",
"Martin",
""
],
[
"Darmawan",
"Andrew",
""
]
] |
2410.23575 | Ryosuke Kadono | Ryosuke Kadono, Takashi U. Ito | Distinguishing Ion Dynamics from Muon Diffusion in Muon Spin Relaxation II -- Extension to Paramagnetic Muons | 10 pages, 13 figures | J. Phys. Soc. Jpn. 94, 064601 (2025) | 10.7566/JPSJ.94.064601 | null | physics.atom-ph cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-sa/4.0/ | We extend the previously published model that distinguishes between the diffusive motion of diamagnetic muons and the dynamics of ions around the muon in matter, and propose a generalized model for {\sl paramagnetic muons} (Mu$^0$s, bound states of a muon and an unpaired electron) observed in non-metallic host materials. The new model distinguishes among the independent motion of unpaired electron associated with Mu$^0$, the self-diffusive motion of Mu$^0$ as single atomic entity, and that of the ions surrounding Mu$^0$, where the muon spin relaxation is induced by dynamical fluctuations of the hyperfine (HF) field exerted from the unpaired electron (e.g., due to spin/charge exchange reaction) and/or that of the nuclear hyperfine (NHF) interactions between the unpaired electron and the surrounding ions. We have applied this model to the muonated radicals (Mu$^0$s in a polaron state) in conducting polymers, and examined the validity of the interpretations claimed in the earlier literature that the spin relaxation is induced by quasi-one dimensional motion of the unpaired electron. The result suggests that experimental support for such a claim is still inadequate and needs to be reexamined, including the possibility of other origins for the fluctuations. It is expected that our model will prove a useful guide for $\mu$SR studies of various local dynamics involving paramagnetic muon states. | [
{
"version": "v1",
"created": "Thu, 31 Oct 2024 02:38:27 GMT"
},
{
"version": "v2",
"created": "Fri, 31 Jan 2025 05:46:06 GMT"
},
{
"version": "v3",
"created": "Thu, 01 May 2025 05:10:48 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kadono",
"Ryosuke",
""
],
[
"Ito",
"Takashi U.",
""
]
] |
2411.00152 | Zahra Aminzare | Pake Melland, Rodica Curtu, Zahra Aminzare | Spike-Adding Mechanisms in a Three-Timescale System: Insights from the FitzHugh-Nagumo Model with Periodic Forcing | 35 pages, 12 figures | null | null | null | math.DS q-bio.NC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we investigate the spike-adding mechanism in a class of three-dimensional fast-slow systems with three distinct timescales, inspired by the FitzHugh-Nagumo (FHN) model driven by periodic input. First, we numerically generate a spike-adding diagram for the FHN model by varying the frequency and amplitude of the input, revealing that as the frequency decreases and the amplitude increases, the number of spikes within each burst grows. We demonstrate that a similar spike-adding structure occurs in the more realistic, periodically forced Morris-Lecar neuronal model. Next, we apply methods from geometric singular perturbation theory to compute critical and super-critical manifolds of the fast-slow system. We use them to characterize the emergence of new burst-spikes in the FHN model, when the periodic forcing resembles a low frequency-band brain rhythm. We then describe how the uncovered spike-adding mechanism defines the boundaries that separate regions with different spike counts in the parameter space. | [
{
"version": "v1",
"created": "Thu, 31 Oct 2024 18:59:14 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:26:14 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Melland",
"Pake",
""
],
[
"Curtu",
"Rodica",
""
],
[
"Aminzare",
"Zahra",
""
]
] |
2411.00387 | Jiaru Zou | Jiaru Zou, Qing Wang, Pratyush Thakur, Nickvash Kani | STEM-POM: Evaluating Language Models Math-Symbol Reasoning in Document Parsing | ACL 2025; NeurIPS Math-AI 2024 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Advances in large language models (LLMs) have spurred research into enhancing their reasoning capabilities, particularly in math-rich STEM (Science, Technology, Engineering, and Mathematics) documents. While LLMs can generate equations or solve math-related queries, their ability to fully understand and interpret abstract mathematical symbols in long, math-rich documents remains limited. In this paper, we introduce STEM-PoM, a comprehensive benchmark dataset designed to evaluate LLMs' reasoning abilities on math symbols within contextual scientific text. The dataset, sourced from real-world ArXiv documents, contains over 2K math symbols classified as main attributes of variables, constants, operators, and unit descriptors, with additional sub-attributes including scalar/vector/matrix for variables and local/global/discipline-specific labels for both constants and operators. Our extensive experiments demonstrate that state-of-the-art LLMs achieve an average accuracy of 20-60% under in-context learning and 50-60% with fine-tuning, highlighting a substantial gap in their ability to classify mathematical symbols. By improving LLMs' mathematical symbol classification, STEM-PoM further enhances models' downstream mathematical reasoning capabilities. The code and data are available at https://github.com/jiaruzouu/STEM-PoM. | [
{
"version": "v1",
"created": "Fri, 01 Nov 2024 06:25:06 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 02:30:26 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zou",
"Jiaru",
""
],
[
"Wang",
"Qing",
""
],
[
"Thakur",
"Pratyush",
""
],
[
"Kani",
"Nickvash",
""
]
] |
2411.01135 | WeiHsiang Liao | WeiHsiang Liao, Yuhta Takida, Yukara Ikemiya, Zhi Zhong, Chieh-Hsin Lai, Giorgio Fabbro, Kazuki Shimada, Keisuke Toyama, Kinwai Cheuk, Marco A. Mart\'inez-Ram\'irez, Shusuke Takahashi, Stefan Uhlich, Taketo Akama, Woosung Choi, Yuichiro Koyama, Yuki Mitsufuji | Music Foundation Model as Generic Booster for Music Downstream Tasks | 41 pages with 14 figures | Published in Transactions on Machine Learning Research (TMLR), 2025 | null | null | cs.SD cs.IR cs.LG eess.AS | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We demonstrate the efficacy of using intermediate representations from a single foundation model to enhance various music downstream tasks. We introduce SoniDo, a music foundation model (MFM) designed to extract hierarchical features from target music samples. By leveraging hierarchical intermediate features, SoniDo constrains the information granularity, leading to improved performance across various downstream tasks including both understanding and generative tasks. We specifically evaluated this approach on representative tasks such as music tagging, music transcription, music source separation, and music mixing. Our results reveal that the features extracted from foundation models provide valuable enhancements in training downstream task models. This highlights the capability of using features extracted from music foundation models as a booster for downstream tasks. Our approach not only benefits existing task-specific models but also supports music downstream tasks constrained by data scarcity. This paves the way for more effective and accessible music processing solutions. | [
{
"version": "v1",
"created": "Sat, 02 Nov 2024 04:44:27 GMT"
},
{
"version": "v2",
"created": "Tue, 05 Nov 2024 08:51:44 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 08:58:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Liao",
"WeiHsiang",
""
],
[
"Takida",
"Yuhta",
""
],
[
"Ikemiya",
"Yukara",
""
],
[
"Zhong",
"Zhi",
""
],
[
"Lai",
"Chieh-Hsin",
""
],
[
"Fabbro",
"Giorgio",
""
],
[
"Shimada",
"Kazuki",
""
],
[
"Toyama",
"Keisuke",
""
],
[
"Cheuk",
"Kinwai",
""
],
[
"Martínez-Ramírez",
"Marco A.",
""
],
[
"Takahashi",
"Shusuke",
""
],
[
"Uhlich",
"Stefan",
""
],
[
"Akama",
"Taketo",
""
],
[
"Choi",
"Woosung",
""
],
[
"Koyama",
"Yuichiro",
""
],
[
"Mitsufuji",
"Yuki",
""
]
] |
2411.01609 | Xiaobin Zhao | Xiaobin Zhao, Pengcheng Liao, Francesco Anna Mele, Ulysse Chabaud, and Quntao Zhuang | Complexity of quantum tomography from genuine non-Gaussian entanglement | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Quantum state tomography, a fundamental tool for quantum physics, usually requires a number of state copies that scale exponentially with the system size, owing to the intricate quantum correlations between subsystems. We show that, in bosonic systems, the nature of correlations indeed fully determines this scaling. Motivated by the Hong-Ou-Mandel effect and Boson-sampling, we define Gaussian-entanglable (GE) states, produced by generalized interference between separable bosonic modes. GE states greatly extend the Gaussian family, encompassing arbitrary separable states, multi-mode Gottesman-Kitaev-Preskill codes, entangled cat states, and Boson-sampling outputs -- resources for error correction and quantum advantage. Nonetheless, we prove that an m-mode pure GE state is learnable with only poly(m) copies, by providing an explicit protocol involving only heterodyne detection and classical post-processing. For states outside GE, we introduce an operational monotone -- the minimum number of ancillary modes required to render them GE -- and prove that it exactly captures the exponential overhead in tomography. As a by-product, we show that deterministic generation of NOON states with N>=3 photons by two-mode interference is impossible. | [
{
"version": "v1",
"created": "Sun, 03 Nov 2024 15:30:06 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 23:15:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhao",
"Xiaobin",
""
],
[
"Liao",
"Pengcheng",
""
],
[
"Mele",
"Francesco Anna",
""
],
[
"Chabaud",
"Ulysse",
""
],
[
"Zhuang",
"Quntao",
""
]
] |
2411.01834 | Guan-Ting Lin | Guan-Ting Lin, Prashanth Gurunath Shivakumar, Aditya Gourav, Yile Gu, Ankur Gandhe, Hung-yi Lee, Ivan Bulyko | Align-SLM: Textless Spoken Language Models with Reinforcement Learning from AI Feedback | Accepted by ACL 2025 | null | null | null | cs.CL eess.AS | http://creativecommons.org/licenses/by-sa/4.0/ | While textless Spoken Language Models (SLMs) have shown potential in end-to-end speech-to-speech modeling, they still lag behind text-based Large Language Models (LLMs) in terms of semantic coherence and relevance. This work introduces the Align-SLM framework, which leverages preference optimization inspired by Reinforcement Learning with AI Feedback (RLAIF) to enhance the semantic understanding of SLMs. Our approach generates multiple speech continuations from a given prompt and uses semantic metrics to create preference data for Direct Preference Optimization (DPO). We evaluate the framework using ZeroSpeech 2021 benchmarks for lexical and syntactic modeling, the spoken version of the StoryCloze dataset for semantic coherence, and other speech generation metrics, including the GPT4-o score and human evaluation. Experimental results show that our method achieves state-of-the-art performance for SLMs on most benchmarks, highlighting the importance of preference optimization to improve the semantics of SLMs. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 06:07:53 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:17:52 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Lin",
"Guan-Ting",
""
],
[
"Shivakumar",
"Prashanth Gurunath",
""
],
[
"Gourav",
"Aditya",
""
],
[
"Gu",
"Yile",
""
],
[
"Gandhe",
"Ankur",
""
],
[
"Lee",
"Hung-yi",
""
],
[
"Bulyko",
"Ivan",
""
]
] |
2411.01878 | Sachitha Bandara | Sachitha C. Bandara, Peter J. Smith, Erfan Khordad, Robin Evans, Rajitha Senanayake | Rician Channel Modelling for Super Wideband MIMO Communications | This paper has been accepted and presented at the IEEE WCNC 2025 | null | null | null | cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent developments in Multiple-Input-Multiple-Output (MIMO) technology include packing a large number of antenna elements in a compact array to access the bandwidth benefits provided by higher mutual coupling (MC). The resulting super-wideband (SW) systems require a circuit-theoretic framework to handle the MC and channel models which span extremely large bands. Hence, in this paper, we make two key contributions. First, we develop a physically-consistent Rician channel model for use with SW systems. Secondly, we express the circuit-theoretic models in terms of a standard MIMO model, so that insights into the effects of antenna layouts, MC, and bandwidth can be made using standard communication theory. For example, we show the bandwidth widening resulting from the new channel model. In addition, we show that MC distorts line-of-sight paths which has beamforming implications. We also highlight the interaction between spatial correlation and MC and show that tight coupling reduces spatial correlations at low frequencies. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 08:11:37 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 00:59:11 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bandara",
"Sachitha C.",
""
],
[
"Smith",
"Peter J.",
""
],
[
"Khordad",
"Erfan",
""
],
[
"Evans",
"Robin",
""
],
[
"Senanayake",
"Rajitha",
""
]
] |
2411.01963 | Imad Ali Shah | Muhammad Waqas Ashraf, Ali Hassan, Imad Ali Shah | V-CAS: A Realtime Vehicle Anti Collision System Using Vision Transformer on Multi-Camera Streams | Accepted at ICMLA 2024 | In 2024 International Conference on Machine Learning and Applications (ICMLA), pp.939-944 | 10.1109/ICMLA61862.2024.00138 | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | This paper introduces a real-time Vehicle Collision Avoidance System (V-CAS) designed to enhance vehicle safety through adaptive braking based on environmental perception. V-CAS leverages the advanced vision-based transformer model RT-DETR, DeepSORT tracking, speed estimation, brake light detection, and an adaptive braking mechanism. It computes a composite collision risk score based on vehicles' relative accelerations, distances, and detected braking actions, using brake light signals and trajectory data from multiple camera streams to improve scene perception. Implemented on the Jetson Orin Nano, V-CAS enables real-time collision risk assessment and proactive mitigation through adaptive braking. A comprehensive training process was conducted on various datasets for comparative analysis, followed by fine-tuning the selected object detection model using transfer learning. The system's effectiveness was rigorously evaluated on the Car Crash Dataset (CCD) from YouTube and through real-time experiments, achieving over 98% accuracy with an average proactive alert time of 1.13 seconds. Results indicate significant improvements in object detection and tracking, enhancing collision avoidance compared to traditional single-camera methods. This research demonstrates the potential of low-cost, multi-camera embedded vision transformer systems to advance automotive safety through enhanced environmental perception and proactive collision avoidance mechanisms. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 10:39:15 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ashraf",
"Muhammad Waqas",
""
],
[
"Hassan",
"Ali",
""
],
[
"Shah",
"Imad Ali",
""
]
] |
2411.02008 | Rujing Xiong | Rujing Xiong, Ke Yin, Jialong Lu, Kai Wan, Tiebin Mi, Robert Caiming Qiu | Flexible Multi-Beam Synthesis and Directional Suppression Through Transmissive RIS | null | null | null | null | cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite extensive research on reconfigurable intelligent surfaces (RISs) in recent years, existing beamforming methods still face significant challenges in achieving flexible and robust beam synthesis, which is an essential capability for a wide range of communication scenarios. This paper introduces a Max-min criterion with nonlinear constraints, leveraging optimization techniques to simultaneously enable flexible multi-beam synthesis and directional suppression using transmissive RIS. Firstly, a realistic model grounded in geometrical optics is introduced to characterize the input/output behaviors of transmissive RISs, effectively bridging the gap between explicit beamforming requirements and practical implementations. Subsequently, a highly efficient algorithm for constrained Max-min optimizations involving quadratic forms is developed. By introducing an auxiliary variable and applying the compensated convexity transform, we successfully reformulate the original non-convex problem and obtain the optimal solution iteratively. This approach is readily applicable to a wide range of constrained Max-min optimization problems. Finally, numerical simulations and prototype experiments are conducted to validate the effectiveness of the proposed framework. The results demonstrate that the proposed algorithm can effectively enhance or selectively suppress signal beams in designated spatial directions, outperforming existing methods in terms of beam control accuracy and robustness. This framework provides valuable insights and references for practical communications applications such as physical layer security and interference mitigation. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 11:51:22 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 10:18:33 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Xiong",
"Rujing",
""
],
[
"Yin",
"Ke",
""
],
[
"Lu",
"Jialong",
""
],
[
"Wan",
"Kai",
""
],
[
"Mi",
"Tiebin",
""
],
[
"Qiu",
"Robert Caiming",
""
]
] |
2411.02138 | Amitai Yacobi | Amitai Yacobi, Ofir Lindenbaum, Uri Shaham | Generalizable and Robust Spectral Method for Multi-view Representation Learning | null | Transactions on Machine Learning Research, 2025 | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multi-view representation learning (MvRL) has garnered substantial attention in recent years, driven by the increasing demand for applications that can effectively process and analyze data from multiple sources. In this context, graph Laplacian-based MvRL methods have demonstrated remarkable success in representing multi-view data. However, these methods often struggle with generalization to new data and face challenges with scalability. Moreover, in many practical scenarios, multi-view data is contaminated by noise or outliers. In such cases, modern deep-learning-based MvRL approaches that rely on alignment or contrastive objectives present degraded performance in downstream tasks, as they may impose incorrect consistency between clear and corrupted data sources. We introduce $\textit{SpecRaGE}$, a novel fusion-based framework that integrates the strengths of graph Laplacian methods with the power of deep learning to overcome these challenges. SpecRage uses neural networks to learn parametric mapping that approximates a joint diagonalization of graph Laplacians. This solution bypasses the need for alignment while enabling generalizable and scalable learning of informative and meaningful representations. Moreover, it incorporates a meta-learning fusion module that dynamically adapts to data quality, ensuring robustness against outliers and noisy views. Our extensive experiments demonstrate that SpecRaGE outperforms state-of-the-art methods, particularly in scenarios with data contamination, paving the way for more reliable and efficient multi-view learning. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 14:51:35 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Jan 2025 19:13:07 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 11:09:59 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Yacobi",
"Amitai",
""
],
[
"Lindenbaum",
"Ofir",
""
],
[
"Shaham",
"Uri",
""
]
] |
2411.02184 | Mouin Ben Ammar | Mou\"in Ben Ammar, David Brellmann, Arturo Mendoza, Antoine Manzanera, Gianni Franchi | Double Descent Meets Out-of-Distribution Detection: Theoretical Insights and Empirical Analysis on the role of model complexity | null | null | null | null | stat.ML cs.AI cs.CV cs.LG math.ST stat.TH | http://creativecommons.org/licenses/by/4.0/ | Out-of-distribution (OOD) detection is essential for ensuring the reliability and safety of machine learning systems. In recent years, it has received increasing attention, particularly through post-hoc detection and training-based methods. In this paper, we focus on post-hoc OOD detection, which enables identifying OOD samples without altering the model's training procedure or objective. Our primary goal is to investigate the relationship between model capacity and its OOD detection performance. Specifically, we aim to answer the following question: Does the Double Descent phenomenon manifest in post-hoc OOD detection? This question is crucial, as it can reveal whether overparameterization, which is already known to benefit generalization, can also enhance OOD detection. Despite the growing interest in these topics by the classic supervised machine learning community, this intersection remains unexplored for OOD detection. We empirically demonstrate that the Double Descent effect does indeed appear in post-hoc OOD detection. Furthermore, we provide theoretical insights to explain why this phenomenon emerges in such setting. Finally, we show that the overparameterized regime does not yield superior results consistently, and we propose a method to identify the optimal regime for OOD detection based on our observations. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 15:39:12 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 09:44:58 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ammar",
"Mouïn Ben",
""
],
[
"Brellmann",
"David",
""
],
[
"Mendoza",
"Arturo",
""
],
[
"Manzanera",
"Antoine",
""
],
[
"Franchi",
"Gianni",
""
]
] |
2411.02549 | Soroosh Shafiee | Daniel Kuhn and Soroosh Shafiee and Wolfram Wiesemann | Distributionally Robust Optimization | null | null | null | null | math.OC cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Distributionally robust optimization (DRO) studies decision problems under uncertainty where the probability distribution governing the uncertain problem parameters is itself uncertain. A key component of any DRO model is its ambiguity set, that is, a family of probability distributions consistent with any available structural or statistical information. DRO seeks decisions that perform best under the worst distribution in the ambiguity set. This worst case criterion is supported by findings in psychology and neuroscience, which indicate that many decision-makers have a low tolerance for distributional ambiguity. DRO is rooted in statistics, operations research and control theory, and recent research has uncovered its deep connections to regularization techniques and adversarial training in machine learning. This survey presents the key findings of the field in a unified and self-contained manner. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 19:32:24 GMT"
},
{
"version": "v2",
"created": "Mon, 28 Apr 2025 22:31:05 GMT"
},
{
"version": "v3",
"created": "Mon, 26 May 2025 19:59:17 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kuhn",
"Daniel",
""
],
[
"Shafiee",
"Soroosh",
""
],
[
"Wiesemann",
"Wolfram",
""
]
] |
2411.02642 | Frederick Nitta | Frederick U. Nitta, Koosha Nassiri Nazif, Eric Pop | Transition Metal Dichalcogenide Solar Cells for Indoor Energy Harvesting | null | null | 10.1016/j.device.2025.100723 | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | With the rapid expansion of the Internet of Things (IoT), efficient and durable energy harvesters for powering IoT devices operating indoors and outdoors are imperative. Promising materials for indoor photovoltaic (PV) technologies include transition metal dichalcogenides (TMDs) such as MoS2, MoSe2, WS2, and WSe2, mainly due to their high absorption coefficients and self-passivated surfaces. Here, we assess the performance of single-junction TMD solar cells under various indoor lighting conditions with a realistic detailed balance model including material-specific optical absorption, as well as radiative, Auger, and defect-assisted Shockley-Read-Hall recombination. We find TMD solar cells could achieve up to 36.5%, 35.6%, 11.2%, and 27.6% power conversion efficiency under fluorescent, LED, halogen, and low-light AM 1.5 G lighting, respectively, at 500 lux. Based on this, TMD solar cells could outperform commercial PV technologies in indoor scenarios, suggesting their viability for future IoT energy solutions. | [
{
"version": "v1",
"created": "Mon, 04 Nov 2024 22:01:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Nitta",
"Frederick U.",
""
],
[
"Nazif",
"Koosha Nassiri",
""
],
[
"Pop",
"Eric",
""
]
] |
2411.03991 | Tatsuya Watanabe | Mathieu Colin and Tatsuya Watanabe | Strong instability of standing waves for $L^2$-supercritical Schr\"odinger-Poisson system with a doping profile | arXiv admin note: substantial text overlap with arXiv:2411.02103; text overlap with arXiv:2409.01842 | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | This paper is devoted to the study of the nonlinear Schr\"odinger-Poisson system with a doping profile. We are interested in the strong instability of standing waves associated with ground state solutions in the $L^2$-supercritical case. The presence of a doping profile causes several difficulties, especially in examining geometric shapes of fibering maps along an $L^2$-invariant scaling curve. Furthermore, the classical approach by Berestycki-Cazenave for the strong instability cannot be applied to our problem due to a remainder term caused by the doping profile. To overcome these difficulties, we establish a new energy inequality associated with the $L^2$-invariant scaling and adopt the strong instability result developed by Fukaya-Ohta(2018). When the doping profile is a characteristic function supported on a bounded smooth domain, some geometric quantities related to the domain, such as the mean curvature, are responsible for the strong instability of standing waves. | [
{
"version": "v1",
"created": "Wed, 06 Nov 2024 15:32:21 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Dec 2024 11:46:03 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 15:01:46 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Colin",
"Mathieu",
""
],
[
"Watanabe",
"Tatsuya",
""
]
] |
2411.04854 | Ramkishor Sharma | Ramkishor Sharma, Axel Brandenburg, Kandaswamy Subramanian, Alexander Vikman | Lattice simulations of axion-U(1) inflation: gravitational waves, magnetic fields, and scalar statistics | 33 pages, 13 figures, Accepted for publication in JCAP | J. Cosmol. Astropart. Phys. 05, 01 (2025) | 10.1088/1475-7516/2025/05/079 | NORDITA-2024-040 | astro-ph.CO hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | We numerically study axion-U(1) inflation, focusing on the regime where the coupling between axions and gauge fields results in significant backreaction from the amplified gauge fields during inflation. These amplified gauge fields not only generate high-frequency gravitational waves (GWs), but also enhance spatial inhomogeneities in the axion field. GWs serve as key probe for constraining the coupling strength between the axion and gauge fields. We find that, when backreaction is important during inflation, the constraints on the coupling strength due to GW overproduction are relaxed compared to previous studies, in which backreaction matters only after inflation. Moreover, our results suggest that the probability density function (PDF) of axion fluctuations tends toward a Gaussian distribution even in cases where gauge field backreaction is important only after inflation. This aligns with previous studies where the same effect was observed for cases with strong backreaction during inflation. This finding can be crucial for future studies of primordial black hole (PBH) formation, which can further constrain the coupling strength. We also calculate the spectrum of the produced magnetic fields in this model and find that their strength is compatible with the observed lower limits. | [
{
"version": "v1",
"created": "Thu, 07 Nov 2024 16:44:39 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Apr 2025 06:33:16 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Sharma",
"Ramkishor",
""
],
[
"Brandenburg",
"Axel",
""
],
[
"Subramanian",
"Kandaswamy",
""
],
[
"Vikman",
"Alexander",
""
]
] |
2411.05671 | Giulia Salatino Mrs | Giulia Salatino, Gianluca Passarelli, Angelo Russomanno, Giuseppe E. Santoro, Procolo Lucignano, Rosario Fazio | Dynamics of edge modes in monitored Su-Schrieffer-Heeger Models | 23 pages, 13 figures | null | null | null | quant-ph cond-mat.mes-hall cond-mat.quant-gas cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | We investigate the effect of dissipation on the dynamics of edge modes in the monitored Su-Schrieffer-Heeger (SSH) model. Our study considers both a linear observable and a nonlinear entanglement measure, namely the two-point correlation function and the Disconnected Entanglement Entropy (DEE), as diagnostic tools. While dissipation inevitably alters the entanglement properties observed in the closed system, statistical analysis of quantum trajectories reveals that by protecting the chain's edges from dissipation, it is possible to recover characteristic features analogous to those found in the unitary limit. This highlights the fundamental role of spatial dissipation patterns in shaping the dynamics of edge modes in monitored systems. | [
{
"version": "v1",
"created": "Fri, 08 Nov 2024 16:22:39 GMT"
},
{
"version": "v2",
"created": "Fri, 15 Nov 2024 08:51:42 GMT"
},
{
"version": "v3",
"created": "Thu, 13 Mar 2025 09:53:21 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 12:21:46 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Salatino",
"Giulia",
""
],
[
"Passarelli",
"Gianluca",
""
],
[
"Russomanno",
"Angelo",
""
],
[
"Santoro",
"Giuseppe E.",
""
],
[
"Lucignano",
"Procolo",
""
],
[
"Fazio",
"Rosario",
""
]
] |
2411.05913 | Konstantinos Tanidis | K. Tanidis, J. Asorey, C. S. Saraf, C. L. Hale, B. Bahr-Kalus, D. Parkinson, S. Camera, R. P. Norris, A. M. Hopkins, M. Bilicki, and N. Gupta | Cross-correlating the EMU Pilot Survey 1 with CMB lensing: Constraints on cosmology and galaxy bias with harmonic-space power spectra | To be submitted to PASA. 22 pages, 9 figures, 4 tables | null | 10.1017/pasa.2025.10034 | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We measured the harmonic-space power spectrum of galaxy clustering auto-correlation from the Evolutionary Map of the Universe Pilot Survey 1 data (EMU PS1) and its cross-correlation with the lensing convergence map of cosmic microwave background (CMB) from Planck Public Release 4 at the linear scale range from $\ell=2$ to 500. We applied two flux density cuts at $0.18$ and $0.4$mJy on the radio galaxies observed at 944MHz and considered two source detection algorithms. We found the auto-correlation measurements from the two algorithms at the 0.18mJy cut to deviate for $\ell\gtrsim250$ due to the different criteria assumed on the source detection and decided to ignore data above this scale. We report a cross-correlation detection of EMU PS1 with CMB lensing at $\sim$5.5$\sigma$, irrespective of flux density cut. In our theoretical modelling we considered two redshift distribution simulation models that yield consistent results, a linear and a non-linear matter power spectrum, and two linear galaxy bias models. That is a constant redshift-independent galaxy bias $b(z)=b_g$ and a constant amplitude galaxy bias $b(z)=b_g/D(z)$. By fixing a cosmology model and considering a non-linear matter power spectrum, we measured a constant galaxy bias at $0.18$mJy ($0.4$mJy) with $b_g=2.32^{+0.41}_{-0.33}$ ($2.18^{+0.17}_{-0.25}$) and a constant amplitude bias with $b_g=1.72^{+0.31}_{-0.21}$ ($1.78^{+0.22}_{-0.15}$). When $\sigma_8$ is a free parameter for the same models at $0.18$mJy ($0.4$mJy) with the constant model we found $\sigma_8=0.68^{+0.16}_{-0.14}$ ($0.82\pm0.10$), while with the constant amplitude model we measured $\sigma_8=0.61^{+0.18}_{-0.20}$ ($0.78^{+0.11}_{-0.09}$), respectively. Our results agree at $1\sigma$ with the measurements from Planck CMB and the weak lensing surveys and also show the potential of cosmology studies with future radio continuum survey data. | [
{
"version": "v1",
"created": "Fri, 08 Nov 2024 19:00:05 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Tanidis",
"K.",
""
],
[
"Asorey",
"J.",
""
],
[
"Saraf",
"C. S.",
""
],
[
"Hale",
"C. L.",
""
],
[
"Bahr-Kalus",
"B.",
""
],
[
"Parkinson",
"D.",
""
],
[
"Camera",
"S.",
""
],
[
"Norris",
"R. P.",
""
],
[
"Hopkins",
"A. M.",
""
],
[
"Bilicki",
"M.",
""
],
[
"Gupta",
"N.",
""
]
] |
2411.06308 | Ezgi Demircan-Tureyen | Ezgi Demircan-Tureyen, Felix Lucka, Tristan van Leeuwen | Exploring Out-of-distribution Detection for Sparse-view Computed Tomography with Diffusion Models | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent works demonstrate the effectiveness of diffusion models as unsupervised solvers for inverse imaging problems. Sparse-view computed tomography (CT) has greatly benefited from these advancements, achieving improved generalization without reliance on measurement parameters. However, this comes at the cost of potential hallucinations, especially when handling out-of-distribution (OOD) data. To ensure reliability, it is essential to study OOD detection for CT reconstruction across both clinical and industrial applications. This need further extends to enabling the OOD detector to function effectively as an anomaly inspection tool. In this paper, we explore the use of a diffusion model, trained to capture the target distribution for CT reconstruction, as an in-distribution prior. Building on recent research, we employ the model to reconstruct partially diffused input images and assess OOD-ness through multiple reconstruction errors. Adapting this approach for sparse-view CT requires redefining the notions of ``input'' and ``reconstruction error''. Here, we use filtered backprojection (FBP) reconstructions as input and investigate various definitions of reconstruction error. Our proof-of-concept experiments on the MNIST dataset highlight both successes and failures, demonstrating the potential and limitations of integrating such an OOD detector into a CT reconstruction system. Our findings suggest that effective OOD detection can be achieved by comparing measurements with forward-projected reconstructions, provided that reconstructions from noisy FBP inputs are conditioned on the measurements. However, conditioning can sometimes lead the OOD detector to inadvertently reconstruct OOD images well. To counter this, we introduce a weighting approach that improves robustness against highly informative OOD measurements, albeit with a trade-off in performance in certain cases. | [
{
"version": "v1",
"created": "Sat, 09 Nov 2024 23:17:42 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 10:46:38 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Demircan-Tureyen",
"Ezgi",
""
],
[
"Lucka",
"Felix",
""
],
[
"van Leeuwen",
"Tristan",
""
]
] |
2411.06875 | Michele Fioretti | Michele Fioretti and Victor Saint-Jean and Simon C. Smith | NGO Activism: Exposure vs. Influence | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We analyze the timing of NGO campaigns to shed light on NGOs' strategies and how they evolve over time. Data from 2,500 campaigns show that NGOs are six times more likely to launch campaigns on their target's Annual General Meeting (AGM) date. Although this strategy increases media exposure and stakeholder scrutiny, resulting in consumer backlash and related shareholder proposals at the following AGM of the targeted firm, it has no impact on current AGM votes. As NGOs build reputational capital, they adjust their timing to influence AGM votes, revealing the trade-offs they face to drive corporate change. | [
{
"version": "v1",
"created": "Mon, 11 Nov 2024 11:14:27 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Nov 2024 13:48:50 GMT"
},
{
"version": "v3",
"created": "Mon, 24 Feb 2025 17:36:12 GMT"
},
{
"version": "v4",
"created": "Sun, 25 May 2025 14:30:18 GMT"
},
{
"version": "v5",
"created": "Tue, 27 May 2025 12:57:57 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Fioretti",
"Michele",
""
],
[
"Saint-Jean",
"Victor",
""
],
[
"Smith",
"Simon C.",
""
]
] |
2411.07115 | Diego Portillo-S\'anchez | Emilio J. Estrada, Juan Manuel M\'arquez, Diego Portillo-S\'anchez and Pablo Roig | Proton-box contribution to $a_{\mu}^{\rm{HLbL}}$ | 19 pages, 4 figures. Version accepted for publication in Phys. Rev. D | Phys. Rev. D 111, 093008 (2025) | 10.1103/PhysRevD.111.093008 | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We analyze the proton$\text{-}$box contribution to the hadronic light$\text{-}$by$\text{-}$light part of the muon's anomalous magnetic moment, which is the first reported baryonic contribution to this piece. We follow the quark$\text{-}$loop analysis, incorporating the relevant data$\text{-}$driven and lattice proton form factors. Although the heavy mass expansion would yield a contribution of $\mathcal{O}(10^{-10})$, the damping of the form factors in the regions where the kernel peaks, explains our finding $a_{\mu}^{\rm{p-box}}=1.82 (7)\times 10^{-12}$, two orders of magnitude smaller than the forthcoming uncertainty on the $a_{\mu}$ measurement and on its Standard Model prediction. | [
{
"version": "v1",
"created": "Mon, 11 Nov 2024 16:43:06 GMT"
},
{
"version": "v2",
"created": "Thu, 08 May 2025 19:57:27 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Estrada",
"Emilio J.",
""
],
[
"Márquez",
"Juan Manuel",
""
],
[
"Portillo-Sánchez",
"Diego",
""
],
[
"Roig",
"Pablo",
""
]
] |
2411.07446 | Cilin Yan | Cilin Yan, Jingyun Wang, Lin Zhang, Ruihui Zhao, Xiaopu Wu, Kai Xiong, Qingsong Liu, Guoliang Kang, Yangyang Kang | Efficient and Accurate Prompt Optimization: the Benefit of Memory in Exemplar-Guided Reflection | ACL 2025 Main | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Automatic prompt engineering aims to enhance the generation quality of large language models (LLMs). Recent works utilize feedbacks generated from erroneous cases to guide the prompt optimization. During inference, they may further retrieve several semantically-related exemplars and concatenate them to the optimized prompts to improve the performance. However, those works only utilize the feedback at the current step, ignoring historical and unseleccted feedbacks which are potentially beneficial. Moreover, the selection of exemplars only considers the general semantic relationship and may not be optimal in terms of task performance and matching with the optimized prompt. In this work, we propose an Exemplar-Guided Reflection with Memory mechanism (ERM) to realize more efficient and accurate prompt optimization. Specifically, we design an exemplar-guided reflection mechanism where the feedback generation is additionally guided by the generated exemplars. We further build two kinds of memory to fully utilize the historical feedback information and support more effective exemplar retrieval. Empirical evaluations show our method surpasses previous state-of-the-arts with less optimization steps, i.e., improving F1 score by 10.1 on LIAR dataset, and reducing half of the optimization steps on ProTeGi. | [
{
"version": "v1",
"created": "Tue, 12 Nov 2024 00:07:29 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 05:08:53 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Yan",
"Cilin",
""
],
[
"Wang",
"Jingyun",
""
],
[
"Zhang",
"Lin",
""
],
[
"Zhao",
"Ruihui",
""
],
[
"Wu",
"Xiaopu",
""
],
[
"Xiong",
"Kai",
""
],
[
"Liu",
"Qingsong",
""
],
[
"Kang",
"Guoliang",
""
],
[
"Kang",
"Yangyang",
""
]
] |
2411.07909 | Paul Voutier | Paul M. Voutier | $n$-defective Lehmer pairs for small $n$: Corrections and Clarifications | Accepted version by International J. Number Theory | null | null | null | math.NT | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We provide some corrections and clarifications of statements in \cite{Ab} and \cite{BHV} regarding elements in Lehmer sequences that have no primitive divisors for small $n$. | [
{
"version": "v1",
"created": "Tue, 12 Nov 2024 16:33:55 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 07:23:47 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Voutier",
"Paul M.",
""
]
] |
2411.08464 | Chao Huang | Chao Huang, Jiahui Chen, Chen Chen, Chunyan Chen, Renjie Su, Shiyu Du, ChenChen, Hongrui Liang, Daojing Lin | A Generation Framework with Strict Constraints for Crystal Materials Design | null | null | null | null | cs.AI cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The design of crystal materials plays a critical role in areas such as new energy development, biomedical engineering, and semiconductors. Recent advances in data-driven methods have enabled the generation of diverse crystal structures. However, most existing approaches still rely on random sampling without strict constraints, requiring multiple post-processing steps to identify stable candidates with the desired physical and chemical properties. In this work, we present a new constrained generation framework that takes multiple constraints as input and enables the generation of crystal structures with specific chemical and properties. In this framework, intermediate constraints, such as symmetry information and composition ratio, are generated by a constraint generator based on large language models (LLMs), which considers the target properties. These constraints are then used by a subsequent crystal structure generator to ensure that the structure generation process is under control. Our method generates crystal structures with a probability of meeting the target properties that is more than twice that of existing approaches. Furthermore, nearly 100% of the generated crystals strictly adhere to predefined chemical composition, eliminating the risks of supply chain during production. | [
{
"version": "v1",
"created": "Wed, 13 Nov 2024 09:36:50 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 06:49:55 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Huang",
"Chao",
""
],
[
"Chen",
"Jiahui",
""
],
[
"Chen",
"Chen",
""
],
[
"Chen",
"Chunyan",
""
],
[
"Su",
"Renjie",
""
],
[
"Du",
"Shiyu",
""
],
[
"ChenChen",
"",
""
],
[
"Liang",
"Hongrui",
""
],
[
"Lin",
"Daojing",
""
]
] |
2411.08810 | Bo Zou | Bo Zou and Allan H. MacDonald | Vortex lattice states of bilayer electron-hole fluids in quantizing magnetic fields | null | null | null | null | cond-mat.mes-hall cond-mat.quant-gas cond-mat.str-el | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We show that the ground state of a weakly charged two-dimensional electron-hole fluid in a strong magnetic field is a broken translation symmetry state with interpenetrating lattices of localized vortices and antivortices in the electron-hole-pair field. The vortices and antivortices carry fractional charges of equal sign but unequal magnitude and have a honeycomb lattice structure that contrasts with the triangular lattices of superconducting electron-electron-pair vortex lattices. We predict that increasing charge density or weakening magnetic field drives a vortex delocalization transition that would be signaled experimentally by an abrupt increase in counterflow transport resistance. | [
{
"version": "v1",
"created": "Wed, 13 Nov 2024 17:46:28 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 05:31:07 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zou",
"Bo",
""
],
[
"MacDonald",
"Allan H.",
""
]
] |
2411.08988 | Stewart Mallory | A. Akintunde, P. Bayati, H. Row, and S.A. Mallory | Single-File Diffusion of Active Brownian Particles | null | J. Chem. Phys. 162, 164902 (2025) | 10.1063/5.0248772 | null | cond-mat.soft cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Single-file diffusion (SFD) is a key mechanism underlying transport phenomena in confined physical and biological systems. In a typical SFD process, microscopic particles are restricted to moving in a narrow channel where they cannot pass one another, resulting in constrained motion and anomalous long-time diffusion. In this study, we use Brownian dynamics simulations and analytical theory to investigate the SFD of athermal active Brownian particles (ABPs), a minimal model of active colloids. Building on prior work (Phys. Rev. E 108, 064601 (2023)), where the kinetic temperature, pressure, and compressibility of the single-file ABP system were derived, we develop an accurate analytical expression for the mean square displacement (MSD) of a tagged particle. We find that the MSD exhibits ballistic behavior at short times, governed by the reduced kinetic temperature of the system. At long times, the characteristic subdiffusive scaling of SFD, $\langle (\Delta x)^2 \rangle \sim t^{1/2}$, is preserved. However, self-propulsion introduces significant changes to the 1D mobility, which we directly relate to the system's compressibility. Furthermore, we demonstrate that the generalized 1D mobility, originally proposed by Kollmann for equilibrium systems (Phys. Rev. Lett. 90, 180602 (2003)), can be extended to active systems with minimal modification. These findings provide a framework for understanding particle transport in active systems and for tuning transport properties at the microscale, particularly in geometries where motion is highly restricted. | [
{
"version": "v1",
"created": "Wed, 13 Nov 2024 19:22:59 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 19:52:51 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Akintunde",
"A.",
""
],
[
"Bayati",
"P.",
""
],
[
"Row",
"H.",
""
],
[
"Mallory",
"S. A.",
""
]
] |
2411.09494 | Frederik Bert Baalbergen MSc. | F.B. Baalbergen and I.E. Zadeh and M.J.A. de Dood | Laser-Induced Relaxation Oscillations in Superconducting Nanobridge Single Photon Detectors | null | null | null | null | cond-mat.supr-con physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We demonstrate novel laser-induced relaxation oscillations in superconducting nanowire single photon detectors (SNSPDs). These oscillations appear when a voltage biased NbTiN nanobridge detector is illuminated with intense pulsed laser light at a repetition rate of $\sim19\text{MHz}$. They differ from the well-known relaxation oscillations by a step-wise increase in frequency and phase locking of the oscillations to the laser pulses. We create a model that incorporates electrical feedback and excludes thermal effects to simulate and explain the origin of the observed laser-induced relaxation oscillations. Qualitative agreement to the experiment is achieved using realistic values for the parameters in the model. | [
{
"version": "v1",
"created": "Thu, 14 Nov 2024 15:01:29 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Feb 2025 14:34:34 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 13:57:39 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Baalbergen",
"F. B.",
""
],
[
"Zadeh",
"I. E.",
""
],
[
"de Dood",
"M. J. A.",
""
]
] |
2411.09628 | Papia Panda | Papia Panda, Priya Mishra, Samiran Roy, Monojit Ghosh and Rukmani Mohanta | Study of large extra dimension and neutrino decay at P2SO experiment | 28 pages, 10 figures, 5 tables | JHEP 05 (2025) 018 | 10.1007/JHEP05(2025)018 | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this study, we explore two intriguing new physics scenarios: the theory of Large Extra Dimensions (LED) and the theory of neutrino decay. We analyze the impact of LED on neutrino oscillations in the contexts of Protvino to Super-ORCA (P2SO), DUNE, and T2HK, with a particular emphasis on P2SO. In contrast, the effects of neutrino decay are examined exclusively in the context of P2SO. For the LED scenario, we find that combining data from P2SO, DUNE, and T2HK can yield tighter constraints than current bounds, but only if all oscillation parameters are measured with high precision. In the case of neutrino decay, P2SO can achieve slightly better bounds compared to ESSnuSB and MOMENT, although its bounds remain weaker than those provided by DUNE and T2HK. Regarding sensitivities to unresolved oscillation parameters, the existence of LED has a minimal impact on the determination of CP violation, mass ordering and octant. However, neutrino decay can significantly influence the sensitivities related to CP violation and octant in a non-trivial manner. | [
{
"version": "v1",
"created": "Thu, 14 Nov 2024 17:55:20 GMT"
},
{
"version": "v2",
"created": "Thu, 09 Jan 2025 10:55:06 GMT"
},
{
"version": "v3",
"created": "Fri, 09 May 2025 19:01:09 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Panda",
"Papia",
""
],
[
"Mishra",
"Priya",
""
],
[
"Roy",
"Samiran",
""
],
[
"Ghosh",
"Monojit",
""
],
[
"Mohanta",
"Rukmani",
""
]
] |
2411.10437 | Reza Kashtiban | Reza Kashtiban and Gavin W. Morley and Mark E. Newton and A T M Anishur Rahman | Nitrogen vacancy center in diamond-based Faraday magnetometer | 7 pages, 5 figures | null | null | null | physics.app-ph quant-ph | http://creativecommons.org/licenses/by/4.0/ | The nitrogen vacancy (NV) center in diamond is a versatile color center used for magnetometry, quantum computing, and quantum communications. In this article, using a single laser beam as a pump and probe, we measure the spin states of the NV center using the Faraday effect and use such measurements to develop a novel magnetic field sensor. Using the spin-state-dependent effect on the left and right circularly polarized light, we probe and confirm the existence of spin-orbit coupling in the NV center at room temperature. The sensitivity of our magnetometer is $350~$nT/$\sqrt{Hz}$, limited by the background produced by the laser trapped inside the diamond. We argue that by using an optical cavity and a high-purity diamond, sensitivities in the femtotesla level can be achieved. | [
{
"version": "v1",
"created": "Fri, 15 Nov 2024 18:56:34 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:02:41 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kashtiban",
"Reza",
""
],
[
"Morley",
"Gavin W.",
""
],
[
"Newton",
"Mark E.",
""
],
[
"Rahman",
"A T M Anishur",
""
]
] |
2411.10774 | Rishabh Upadhyay | Rishabh Upadhyay, Bayan Karimi, Diego Subero, Christoforus Dimas Satrya, Joonas T. Peltonen, Yu-Cheng Chang, and Jukka P. Pekola | Strong-coupling quantum thermodynamics using a superconducting flux qubit | 18 pages, 12 Figures | null | null | null | quant-ph cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Thermodynamics in quantum circuits aims to find improved functionalities of thermal machines, highlight fundamental phenomena peculiar to quantum nature in thermodynamics, and point out limitations in quantum information processing due to coupling of the system to its environment. An important aspect to achieve some of these goals is the regime of strong coupling that has remained until now a domain of theoretical works only. Our aim is to demonstrate strong coupling features in heat transport using a superconducting flux qubit, which is capable of reaching strong to deep-ultra strong coupling regimes, as shown in previous studies. Here, we show experimental evidence of strong coupling by observing a hybridized state of the qubit with two cavities coupled to it, leading to a triplet-like thermal transport via this combined system around the minimum energy of the qubit, at power levels of tens of femtowatts, exceeding by an order of magnitude those in earlier experiments. We also demonstrate close to 100% on-off switching ratio of heat current mediated by photons by applying magnetic flux to the qubit. Our experiment opens a way towards testing debated questions in strong coupling thermodynamics such as what heat in this regime is. We also present a theoretical model that aligns with our experimental findings and explains the mechanism behind heat transport in our device. Furthermore, our experiment opens new possibilities for quantum thermodynamics, aiming to realize true quantum heat engines and refrigerators with enhanced power and efficiency, by leveraging ultra-strong coupling between the system and its environment in future experiments. | [
{
"version": "v1",
"created": "Sat, 16 Nov 2024 11:20:05 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 08:34:28 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Upadhyay",
"Rishabh",
""
],
[
"Karimi",
"Bayan",
""
],
[
"Subero",
"Diego",
""
],
[
"Satrya",
"Christoforus Dimas",
""
],
[
"Peltonen",
"Joonas T.",
""
],
[
"Chang",
"Yu-Cheng",
""
],
[
"Pekola",
"Jukka P.",
""
]
] |
2411.11367 | Vladimir Bordo | V.G. Bordo | Proposal for an electrically controlled terahertz micro-oscillator | 4 pages | AIP Advances, vol. 15, 055032 (2025) | 10.1063/5.0256137 | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A source which generates sustained terahertz electromagnetic radiation (oscillator) controllable by an applied voltage is proposed. The structure consists of a nanocomposite slab containing metal nanorods enclosed between two graphene sheets which are electrostatically gated. The principle of its operation is based on a positive feedback which the nanorods polarization experiences from the radiation reflected by the graphene mirrors. The proposed approach can provide a microscopic source of monochromatic polarized terahertz radiation of power sufficient for biomedical and security applications. | [
{
"version": "v1",
"created": "Mon, 18 Nov 2024 08:25:21 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bordo",
"V. G.",
""
]
] |
2411.11830 | Sara Ortega-Martinez | Sara Ortega-Martinez, Sergio Contreras, Raul E. Angulo, Jonas Chaves-Montero | Investigating the galaxy-halo connection of DESI Emission-Line Galaxies with SHAMe-SF | 23 pages, 17 figures. To be submitted to A&A | A&A 697, A226 (2025) | 10.1051/0004-6361/202453086 | null | astro-ph.CO astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | The Dark Energy Spectroscopic Instrument (DESI) survey is mapping the large-scale distribution of millions of Emission Line Galaxies (ELGs) over vast cosmic volumes to measure the growth history of the Universe. However, compared to Luminous Red Galaxies (LRGs), very little is known about the connection of ELGs with the underlying matter field. In this paper, we employ a novel theoretical model, SHAMe-SF, to infer the connection between ELGs and their host dark matter subhaloes. SHAMe-SF is a version of subhalo abundance matching that incorporates prescriptions for multiple processes, including star formation, tidal stripping, environmental correlations, and quenching. We analyse the public measurements of the projected and redshift-space ELGs correlation functions at $z=1.0$ and $z=1.3$ from DESI One Percent data release, which we fit over a broad range of scales $r \in [0.1, 30]/h^{-1}$Mpc to within the statistical uncertainties of the data. We also validate the inference pipeline using two mock DESI ELG catalogues built from hydrodynamical (TNG300) and semi-analytical galaxy formation models (\texttt{L-Galaxies}). SHAMe-SF is able to reproduce the clustering of DESI-ELGs and the mock DESI samples within statistical uncertainties. We infer that DESI ELGs typically reside in haloes of $\sim 10^{11.8}h^{-1}$M$_{\odot}$ when they are central, and $\sim 10^{12.5}h^{-1}$M$_{\odot}$ when they are a satellite, which occurs in $\sim$30 \% of the cases. In addition, compared to the distribution of dark matter within halos, satellite ELGs preferentially reside both in the outskirts and inside haloes, and have a net infall velocity towards the centre. Finally, our results show evidence of assembly bias and conformity. | [
{
"version": "v1",
"created": "Mon, 18 Nov 2024 18:48:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ortega-Martinez",
"Sara",
""
],
[
"Contreras",
"Sergio",
""
],
[
"Angulo",
"Raul E.",
""
],
[
"Chaves-Montero",
"Jonas",
""
]
] |
2411.12544 | Weizhao Chen | Weizhao Chen, Yu Zhang, Yi Liu, and Zhe Yuan | Symmetry-forbidden intraband transitions leading to ultralow Gilbert damping in van der Waals ferromagnets | 6 pages, 5 figures, van der Waals ferromagnet, Gilbert damping, mirror symmetry, first-principles calculation | null | null | null | cond-mat.mtrl-sci cond-mat.mes-hall | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Based upon first-principles calculations, we report ultralow Gilbert damping in two-dimensional (2D) van der Waals (vdW) ferromagnets. The low damping occurs at low temperature because mirror symmetry prohibits intraband transitions. The monotonic dependence on the electronic scattering rate suggests the absent lower limit, in contrast to conventional ferromagnetic materials. Breaking mirror symmetry through magnetization rotation, layer stacking, or structural phase transition significantly increases damping by enabling intraband transitions. Topological nodal lines, also protected by mirror symmetry, contribute substantially to interband-transition-mediated damping, which can be tuned by adjusting the Fermi level. Our findings elucidate the unique characteristics of Gilbert damping in 2D vdW ferromagnets, providing valuable insights for designing low-dimensional spintronic devices with high energy efficiency. | [
{
"version": "v1",
"created": "Tue, 19 Nov 2024 14:50:07 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 07:56:28 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chen",
"Weizhao",
""
],
[
"Zhang",
"Yu",
""
],
[
"Liu",
"Yi",
""
],
[
"Yuan",
"Zhe",
""
]
] |
2411.12719 | Praveen S V | Praveen Srinivasa Varadhan, Amogh Gulati, Ashwin Sankar, Srija Anand, Anirudh Gupta, Anirudh Mukherjee, Shiva Kumar Marepally, Ankur Bhatia, Saloni Jaju, Suvrat Bhooshan, Mitesh M. Khapra | Rethinking MUSHRA: Addressing Modern Challenges in Text-to-Speech Evaluation | Accepted in TMLR | null | null | null | cs.CL cs.LG cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | Despite rapid advancements in TTS models, a consistent and robust human evaluation framework is still lacking. For example, MOS tests fail to differentiate between similar models, and CMOS's pairwise comparisons are time-intensive. The MUSHRA test is a promising alternative for evaluating multiple TTS systems simultaneously, but in this work we show that its reliance on matching human reference speech unduly penalises the scores of modern TTS systems that can exceed human speech quality. More specifically, we conduct a comprehensive assessment of the MUSHRA test, focusing on its sensitivity to factors such as rater variability, listener fatigue, and reference bias. Based on our extensive evaluation involving 492 human listeners across Hindi and Tamil we identify two primary shortcomings: (i) reference-matching bias, where raters are unduly influenced by the human reference, and (ii) judgement ambiguity, arising from a lack of clear fine-grained guidelines. To address these issues, we propose two refined variants of the MUSHRA test. The first variant enables fairer ratings for synthesized samples that surpass human reference quality. The second variant reduces ambiguity, as indicated by the relatively lower variance across raters. By combining these approaches, we achieve both more reliable and more fine-grained assessments. We also release MANGO, a massive dataset of 246,000 human ratings, the first-of-its-kind collection for Indian languages, aiding in analyzing human preferences and developing automatic metrics for evaluating TTS systems. | [
{
"version": "v1",
"created": "Tue, 19 Nov 2024 18:37:45 GMT"
},
{
"version": "v2",
"created": "Sun, 15 Dec 2024 20:38:16 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:40:41 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Varadhan",
"Praveen Srinivasa",
""
],
[
"Gulati",
"Amogh",
""
],
[
"Sankar",
"Ashwin",
""
],
[
"Anand",
"Srija",
""
],
[
"Gupta",
"Anirudh",
""
],
[
"Mukherjee",
"Anirudh",
""
],
[
"Marepally",
"Shiva Kumar",
""
],
[
"Bhatia",
"Ankur",
""
],
[
"Jaju",
"Saloni",
""
],
[
"Bhooshan",
"Suvrat",
""
],
[
"Khapra",
"Mitesh M.",
""
]
] |
2411.13112 | Xianda Guo | Xianda Guo and Ruijun Zhang and Yiqun Duan and Yuhang He and Dujun Nie and Wenke Huang and Chenming Zhang and Shuai Liu and Hao Zhao and Long Chen | SURDS: Benchmarking Spatial Understanding and Reasoning in Driving Scenarios with Vision Language Models | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate spatial reasoning in outdoor environments - covering geometry, object pose, and inter-object relationships - is fundamental to downstream tasks such as mapping, motion forecasting, and high-level planning in autonomous driving. We introduce SURDS, a large-scale benchmark designed to systematically evaluate the spatial reasoning capabilities of vision language models (VLMs). Built on the nuScenes dataset, SURDS comprises 41,080 vision-question-answer training instances and 9,250 evaluation samples, spanning six spatial categories: orientation, depth estimation, pixel-level localization, pairwise distance, lateral ordering, and front-behind relations. We benchmark leading general-purpose VLMs, including GPT, Gemini, and Qwen, revealing persistent limitations in fine-grained spatial understanding. To address these deficiencies, we go beyond static evaluation and explore whether alignment techniques can improve spatial reasoning performance. Specifically, we propose a reinforcement learning-based alignment scheme leveraging spatially grounded reward signals - capturing both perception-level accuracy (location) and reasoning consistency (logic). We further incorporate final-answer correctness and output-format rewards to guide fine-grained policy adaptation. Our GRPO-aligned variant achieves an overall score of 40.80 in the SURDS benchmark. Notably, it outperforms proprietary systems such as GPT-4o (13.30) and Gemini-2.0-flash (35.71). To our best knowledge, this is the first study to demonstrate that reinforcement learning-based alignment can significantly and consistently enhance the spatial reasoning capabilities of VLMs in real-world driving contexts. We release the SURDS benchmark, evaluation toolkit, and GRPO alignment code through: https://github.com/XiandaGuo/Drive-MLLM. | [
{
"version": "v1",
"created": "Wed, 20 Nov 2024 08:14:01 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Nov 2024 07:24:04 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 14:49:49 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Guo",
"Xianda",
""
],
[
"Zhang",
"Ruijun",
""
],
[
"Duan",
"Yiqun",
""
],
[
"He",
"Yuhang",
""
],
[
"Nie",
"Dujun",
""
],
[
"Huang",
"Wenke",
""
],
[
"Zhang",
"Chenming",
""
],
[
"Liu",
"Shuai",
""
],
[
"Zhao",
"Hao",
""
],
[
"Chen",
"Long",
""
]
] |
2411.13940 | Long Zhang | Hongting Hou, Long Zhang | Preparation and observation of anomalous counterpropagating edge states in a periodically driven optical Raman lattice | 15 pages, 11 figures | Phys. Rev. A 111, 053318 (2025) | 10.1103/PhysRevA.111.053318 | null | cond-mat.quant-gas quant-ph | http://creativecommons.org/licenses/by/4.0/ | Motivated by the recent observation of real-space edge modes with ultracold atoms [Braun et al., Nat. Phys. 20, 1306 (2024)], we investigate the preparation and detection of anomalous counterpropagating edge states -- a defining feature of the anomalous Floquet valley-Hall (AFVH) phase -- in a two-dimensional periodically driven optical Raman lattice. Modeling the atomic cloud with a Gaussian wave packet state, we explore, both analytically and numerically, how the population of edge modes depends on the initial-state parameters. In particular, we reveal that, in addition to the internal spin state, the initial momenta parallel and perpendicular to the boundary play essential roles: they independently control the selective population of edge states across distinct momenta and within separate quasienergy gaps. Furthermore, we examine the wave-packet dynamics of counterpropagating edge states and demonstrate that their characteristic motion is robust against long-range disorder. These results establish a theoretical framework for future experimental explorations of the AFVH phase and topological phenomena associated with its unique edge modes. | [
{
"version": "v1",
"created": "Thu, 21 Nov 2024 08:47:29 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Nov 2024 17:40:41 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 14:47:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Hou",
"Hongting",
""
],
[
"Zhang",
"Long",
""
]
] |
2411.14055 | Hexuan Deng | Hexuan Deng, Wenxiang Jiao, Xuebo Liu, Jing Li, Min Zhang, Zhaopeng Tu | DRPruning: Efficient Large Language Model Pruning through Distributionally Robust Optimization | Accepted by ACL 2025 Main Conference | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) deliver impressive results but face challenges from increasing model sizes and computational costs. Structured pruning reduces model size and speeds up inference but often causes uneven degradation across domains, leading to biased performance. To address this, we propose DRPruning, a method that dynamically adjusts the data distribution during training to restore balanced performance across heterogeneous and multi-tasking data. Experiments in monolingual and multilingual settings show that DRPruning surpasses similarly sized models in both pruning and continued pretraining over perplexity, downstream tasks, and instruction tuning. Further analysis demonstrates the robustness of DRPruning towards various domains and distribution shifts. Furthermore, DRPruning can determine optimal reference losses and data ratios automatically, suggesting potential for broader applications. Code and scripts are available at https://github.com/hexuandeng/DRPruning. | [
{
"version": "v1",
"created": "Thu, 21 Nov 2024 12:02:39 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 02:52:03 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Deng",
"Hexuan",
""
],
[
"Jiao",
"Wenxiang",
""
],
[
"Liu",
"Xuebo",
""
],
[
"Li",
"Jing",
""
],
[
"Zhang",
"Min",
""
],
[
"Tu",
"Zhaopeng",
""
]
] |
2411.14098 | Shao Hung Chung | Shao-Hung Chung, Wei Chen, H. H. Jen | Strongly Confined Atomic Excitation Localization in a Weakly-Driven Atom-Waveguide Interface | 8 figures | J. Phys. B: At. Mol. Opt. Phys. 58, 105502 (2025) | 10.1088/1361-6455/add77a | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An atomic array coupled to a photonic crystal waveguide forms a strongly coupled quantum interface, exhibiting various intriguing collective features of quantum dynamics. Here we consider a homogeneous atomic array and theoretically investigate its steady-state distribution when the incident fields drive the atoms from both sides at asymmetric angles. This effectively creates an interface shared by two zones of atoms under different driving angles. This setup introduces a competition between photon-mediated dipole-dipole interactions and the directionality of coupling, while differences of the travelling phases from the incident angles further influence the overall steady-state behavior. Under this asymmetric driving scheme, the presence of strongly confined localization can be identified, where localization can occur either at the interface or at one of edges. Additionally, we examine the size effect on the atomic localization, deriving an empirical formula to predict parameter regimes that favor interfaced localization. We also consider a defect-driving scheme, where a third zone is created by undriven atoms under symmetric travelling phases. This results in strongly confined single-site excitation localization, which can be explained through analytical solutions under the reciprocal coupling. Finally, we propose several methods for precise control of multiple single-site localizations under the defect-driving scheme. Our results provide insights into driven-dissipative quantum systems with nonreciprocal couplings and pave the way for quantum simulation of exotic many-body states relevant to quantum information applications. | [
{
"version": "v1",
"created": "Thu, 21 Nov 2024 13:04:22 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chung",
"Shao-Hung",
""
],
[
"Chen",
"Wei",
""
],
[
"Jen",
"H. H.",
""
]
] |
2411.14419 | Benoit Semelin | Benoit Semelin, Romain M\'eriot, Ashutosh Mishra and David Cornu | Combining summary statistics with simulation-based inference for the 21 cm signal from the Epoch of Reionization | 11 pages, submitted to A&A | A&A 698, A35 (2025) | 10.1051/0004-6361/202453115 | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The 21 cm signal from the Epoch of Reionization will be observed with the up-coming Square Kilometer Array (SKA). SKA should yield a full tomography of the signal which opens the possibility to explore its non-Gaussian properties. How can we extract the maximum information from the tomography and derive the tightest constraint on the signal? In this work, instead of looking for the most informative summary statistics, we investigate how to combine the information from two sets of summary statistics using simulation-based inference. To this purpose, we train Neural Density Estimators (NDE) to fit the implicit likelihood of our model, the LICORICE code, using the Loreli II database. We train three different NDEs: one to perform Bayesian inference on the power spectrum, one to do it on the linear moments of the Pixel Distribution Function (PDF) and one to work with the combination of the two. We perform $\sim 900$ inferences at different points in our parameter space and use them to assess both the validity of our posteriors with Simulation-based Calibration (SBC) and the typical gain obtained by combining summary statistics. We find that our posteriors are biased by no more than $\sim 20 \%$ of their standard deviation and under-confident by no more than $\sim 15 \%$. Then, we establish that combining summary statistics produces a contraction of the 4-D volume of the posterior (derived from the generalized variance) in 91.5 % of our cases, and in 70 to 80 % of the cases for the marginalized 1-D posteriors. The median volume variation is a contraction of a factor of a few for the 4D posteriors and a contraction of 20 to 30 % in the case of the marginalized 1D posteriors. This shows that our approach is a possible alternative to looking for sufficient statistics in the theoretical sense. | [
{
"version": "v1",
"created": "Thu, 21 Nov 2024 18:52:29 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Semelin",
"Benoit",
""
],
[
"Mériot",
"Romain",
""
],
[
"Mishra",
"Ashutosh",
""
],
[
"Cornu",
"David",
""
]
] |
2411.15114 | Neev Parikh | Hjalmar Wijk, Tao Lin, Joel Becker, Sami Jawhar, Neev Parikh, Thomas Broadley, Lawrence Chan, Michael Chen, Josh Clymer, Jai Dhyani, Elena Ericheva, Katharyn Garcia, Brian Goodrich, Nikola Jurkovic, Holden Karnofsky, Megan Kinniment, Aron Lajko, Seraphina Nix, Lucas Sato, William Saunders, Maksym Taran, Ben West, Elizabeth Barnes | RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Frontier AI safety policies highlight automation of AI research and development (R&D) by AI agents as an important capability to anticipate. However, there exist few evaluations for AI R&D capabilities, and none that are highly realistic and have a direct comparison to human performance. We introduce RE-Bench (Research Engineering Benchmark, v1), which consists of 7 challenging, open-ended ML research engineering environments and data from 71 8-hour attempts by 61 distinct human experts. We confirm that our experts make progress in the environments given 8 hours, with 82% of expert attempts achieving a non-zero score and 24% matching or exceeding our strong reference solutions. We compare humans to several public frontier models through best-of-k with varying time budgets and agent designs, and find that the best AI agents achieve a score 4x higher than human experts when both are given a total time budget of 2 hours per environment. However, humans currently display better returns to increasing time budgets, narrowly exceeding the top AI agent scores given an 8-hour budget, and achieving 2x the score of the top AI agent when both are given 32 total hours (across different attempts). Qualitatively, we find that modern AI agents possess significant expertise in many ML topics -- e.g. an agent wrote a faster custom Triton kernel than any of our human experts' -- and can generate and test solutions over ten times faster than humans, at much lower cost. We open-source the evaluation environments, human expert data, analysis code and agent trajectories to facilitate future research. | [
{
"version": "v1",
"created": "Fri, 22 Nov 2024 18:30:46 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 03:32:23 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Wijk",
"Hjalmar",
""
],
[
"Lin",
"Tao",
""
],
[
"Becker",
"Joel",
""
],
[
"Jawhar",
"Sami",
""
],
[
"Parikh",
"Neev",
""
],
[
"Broadley",
"Thomas",
""
],
[
"Chan",
"Lawrence",
""
],
[
"Chen",
"Michael",
""
],
[
"Clymer",
"Josh",
""
],
[
"Dhyani",
"Jai",
""
],
[
"Ericheva",
"Elena",
""
],
[
"Garcia",
"Katharyn",
""
],
[
"Goodrich",
"Brian",
""
],
[
"Jurkovic",
"Nikola",
""
],
[
"Karnofsky",
"Holden",
""
],
[
"Kinniment",
"Megan",
""
],
[
"Lajko",
"Aron",
""
],
[
"Nix",
"Seraphina",
""
],
[
"Sato",
"Lucas",
""
],
[
"Saunders",
"William",
""
],
[
"Taran",
"Maksym",
""
],
[
"West",
"Ben",
""
],
[
"Barnes",
"Elizabeth",
""
]
] |
2411.15119 | Rahul Kumar | Rahul Kumar Walia, Prashant Kocherlakota, Dominic O. Chang, and Kiana Salehi | Spacetime Measurements with the Photon Ring | 15 pages and 8 figures. Matches with the published version | Phys. Rev. D 111, 104074, 10, (2025) | 10.1103/PhysRevD.111.104074 | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | We explore the universal symmetries of the black hole photon ring in a wide range of non-Kerr spacetimes, including the Kerr-Newman, Kerr-Sen, Kerr-Bardeen, and Kerr-Hayward metrics. The demagnification exponent ($\gamma$) controls the size and flux scaling of higher-order images, which appear in the photon ring, the time delay ($\tau$) determines the timing of their appearance, and the rotation parameter ($\delta$) relates their relative orientations on the image plane. Our investigation reveals that these critical parameters respond distinctly to variations in black hole spin, generalized charge, and observer inclination, establishing them as complementary probes of spacetime geometry: $\gamma$ is predominantly influenced by charge and spin, $\tau$ is strongly affected by inclination, especially for near-extremal black holes, and $\delta$ is highly sensitive to spin. Notably, we find that the time delay provides an independent constraint on shadow size for polar observers, while the rotation parameter facilitates metric-independent spin measurements. Specifically, for Kerr black holes, the total variation in $\gamma$, $\tau$, and $\delta$ across all possible inclinations and spins is $\lesssim 20\%$, $\lesssim 10\%$, and $\lesssim 60\%$, respectively. By contrast, the Kerr shadow radius varies by only $\lesssim 8\%$. A future joint measurement of these critical parameters -- along with the black hole shadow size -- will enable precise spacetime characterization, including measurements of the spin, inclination, and generalized charge. | [
{
"version": "v1",
"created": "Fri, 22 Nov 2024 18:36:14 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Nov 2024 04:46:56 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 15:25:20 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Walia",
"Rahul Kumar",
""
],
[
"Kocherlakota",
"Prashant",
""
],
[
"Chang",
"Dominic O.",
""
],
[
"Salehi",
"Kiana",
""
]
] |
2411.15626 | Filip Ilievski | Filip Ilievski, Barbara Hammer, Frank van Harmelen, Benjamin Paassen, Sascha Saralajew, Ute Schmid, Michael Biehl, Marianna Bolognesi, Xin Luna Dong, Kiril Gashteovski, Pascal Hitzler, Giuseppe Marra, Pasquale Minervini, Martin Mundt, Axel-Cyrille Ngonga Ngomo, Alessandro Oltramari, Gabriella Pasi, Zeynep G. Saribatur, Luciano Serafini, John Shawe-Taylor, Vered Shwartz, Gabriella Skitalinskaya, Clemens Stachl, Gido M. van de Ven and Thomas Villmann | Aligning Generalisation Between Humans and Machines | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Recent advances in AI -- including generative approaches -- have resulted in technology that can support humans in scientific discovery and forming decisions, but may also disrupt democracies and target individuals. The responsible use of AI and its participation in human-AI teams increasingly shows the need for AI alignment, that is, to make AI systems act according to our preferences. A crucial yet often overlooked aspect of these interactions is the different ways in which humans and machines generalise. In cognitive science, human generalisation commonly involves abstraction and concept learning. In contrast, AI generalisation encompasses out-of-domain generalisation in machine learning, rule-based reasoning in symbolic AI, and abstraction in neurosymbolic AI. In this perspective paper, we combine insights from AI and cognitive science to identify key commonalities and differences across three dimensions: notions of, methods for, and evaluation of generalisation. We map the different conceptualisations of generalisation in AI and cognitive science along these three dimensions and consider their role for alignment in human-AI teaming. This results in interdisciplinary challenges across AI and cognitive science that must be tackled to provide a foundation for effective and cognitively supported alignment in human-AI teaming scenarios. | [
{
"version": "v1",
"created": "Sat, 23 Nov 2024 18:36:07 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:44:21 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ilievski",
"Filip",
""
],
[
"Hammer",
"Barbara",
""
],
[
"van Harmelen",
"Frank",
""
],
[
"Paassen",
"Benjamin",
""
],
[
"Saralajew",
"Sascha",
""
],
[
"Schmid",
"Ute",
""
],
[
"Biehl",
"Michael",
""
],
[
"Bolognesi",
"Marianna",
""
],
[
"Dong",
"Xin Luna",
""
],
[
"Gashteovski",
"Kiril",
""
],
[
"Hitzler",
"Pascal",
""
],
[
"Marra",
"Giuseppe",
""
],
[
"Minervini",
"Pasquale",
""
],
[
"Mundt",
"Martin",
""
],
[
"Ngomo",
"Axel-Cyrille Ngonga",
""
],
[
"Oltramari",
"Alessandro",
""
],
[
"Pasi",
"Gabriella",
""
],
[
"Saribatur",
"Zeynep G.",
""
],
[
"Serafini",
"Luciano",
""
],
[
"Shawe-Taylor",
"John",
""
],
[
"Shwartz",
"Vered",
""
],
[
"Skitalinskaya",
"Gabriella",
""
],
[
"Stachl",
"Clemens",
""
],
[
"van de Ven",
"Gido M.",
""
],
[
"Villmann",
"Thomas",
""
]
] |
2411.16064 | Peihua Deng | Peihua Deng, Jiehua Zhang, Xichun Sheng, Chenggang Yan, Yaoqi Sun, Ying Fu, Liang Li | Multi-Granularity Class Prototype Topology Distillation for Class-Incremental Source-Free Unsupervised Domain Adaptation | Accepted by CVPR 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper explores the Class-Incremental Source-Free Unsupervised Domain Adaptation (CI-SFUDA) problem, where the unlabeled target data come incrementally without access to labeled source instances. This problem poses two challenges, the interference of similar source-class knowledge in target-class representation learning and the shocks of new target knowledge to old ones. To address them, we propose the Multi-Granularity Class Prototype Topology Distillation (GROTO) algorithm, which effectively transfers the source knowledge to the class-incremental target domain. Concretely, we design the multi-granularity class prototype self-organization module and the prototype topology distillation module. First, we mine the positive classes by modeling accumulation distributions. Next, we introduce multi-granularity class prototypes to generate reliable pseudo-labels, and exploit them to promote the positive-class target feature self-organization. Second, the positive-class prototypes are leveraged to construct the topological structures of source and target feature spaces. Then, we perform the topology distillation to continually mitigate the shocks of new target knowledge to old ones. Extensive experiments demonstrate that our proposed method achieves state-of-the-art performance on three public datasets. Code is available at https://github.com/dengpeihua/GROTO. | [
{
"version": "v1",
"created": "Mon, 25 Nov 2024 03:28:09 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Mar 2025 12:35:16 GMT"
},
{
"version": "v3",
"created": "Tue, 18 Mar 2025 08:34:36 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 12:46:03 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Deng",
"Peihua",
""
],
[
"Zhang",
"Jiehua",
""
],
[
"Sheng",
"Xichun",
""
],
[
"Yan",
"Chenggang",
""
],
[
"Sun",
"Yaoqi",
""
],
[
"Fu",
"Ying",
""
],
[
"Li",
"Liang",
""
]
] |
2411.16795 | Yiran Wang | Yiran Wang, Willem Meijer, Jos\'e Antonio Hern\'andez L\'opez, Ulf Nilsson, D\'aniel Varr\'o | Why do Machine Learning Notebooks Crash? An Empirical Study on Public Python Jupyter Notebooks | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Jupyter notebooks have become central in data science, integrating code, text and output in a flexible environment. With the rise of machine learning (ML), notebooks are increasingly used for prototyping and data analysis. However, due to their dependence on complex ML libraries and the flexible notebook semantics that allow cells to be run in any order, notebooks are susceptible to software bugs that may lead to program crashes. This paper presents a comprehensive empirical study focusing on crashes in publicly available Python ML notebooks. We collect 64,031 notebooks containing 92,542 crashes from GitHub and Kaggle, and manually analyze a sample of 746 crashes across various aspects, including crash types and root causes. Our analysis identifies unique ML-specific crash types, such as tensor shape mismatches and dataset value errors that violate API constraints. Additionally, we highlight unique root causes tied to notebook semantics, including out-of-order execution and residual errors from previous cells, which have been largely overlooked in prior research. Furthermore, we identify the most error-prone ML libraries, and analyze crash distribution across ML pipeline stages. We find that over 40% of crashes stem from API misuse and notebook-specific issues. Crashes frequently occur when using ML libraries like TensorFlow/Keras and Torch. Additionally, over 70% of the crashes occur during data preparation, model training, and evaluation or prediction stages of the ML pipeline, while data visualization errors tend to be unique to ML notebooks. | [
{
"version": "v1",
"created": "Mon, 25 Nov 2024 09:33:08 GMT"
},
{
"version": "v2",
"created": "Sun, 27 Apr 2025 09:03:02 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 07:42:09 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Wang",
"Yiran",
""
],
[
"Meijer",
"Willem",
""
],
[
"López",
"José Antonio Hernández",
""
],
[
"Nilsson",
"Ulf",
""
],
[
"Varró",
"Dániel",
""
]
] |
2411.16876 | Yetli Rosas-Guevara | Yetli Rosas-Guevara, Silvia Bonoli, Ewald Puchwein, Massimo Dotti and Sergio Contreras | Galaxy formation physics behind bar formation: A view from cosmological hydrodynamical simulations | 22 pages, 11 figures, Accepted to A&A. Changes in section 4, Abstract and Discussion | A&A 698, A20 (2025) | 10.1051/0004-6361/202453160 | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a suite of zoom-in cosmological simulations of Milky Way-like galaxies with a prominent disc component and a strong bar in their centre, based on a subsample of barred galaxies from the TNG50 magneto-hydrodynamic simulation. We modify the physical models that regulate star formation, namely, supernova feedback and black hole quasar feedback, to examine how they affect the disc and bar formation. We find that, independently of the feedback prescriptions, all galaxies show a similar morphology, which is dominant in comparison with the bulge mass. The black hole quasar feedback models used in this study do not affect bar formation, although they can affect the bar strength and length. The energy released by the supernovae causes a delay in the time of bar formation and, in models with the strongest feedback, galaxies form stable discs against bar formation. This could be understood since supernova feedback influences disc and bulge assembly, resulting in discs with lower mass content, radial velocity dispersion and larger size as the supernova feedback strength increases. We study disc stability using three bar instability criteria proposed in the literature. We find that galaxies with varied supernovae and black hole quasar feedback satisfy these criteria at the moment of bar formation, except in extreme cases where the galaxy lacks or has weak supernova feedback. In these models, two of three criteria fail to forecast the existence (or absence) of a bar, probably because they do not account for the influence of a massive and compact bulge. Our findings provide insights into the physical processes behind bar formation and highlight the importance of additional conditions, other than a massive and compact disc that promote bar formation. | [
{
"version": "v1",
"created": "Mon, 25 Nov 2024 19:16:25 GMT"
},
{
"version": "v2",
"created": "Mon, 07 Apr 2025 09:04:35 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Rosas-Guevara",
"Yetli",
""
],
[
"Bonoli",
"Silvia",
""
],
[
"Puchwein",
"Ewald",
""
],
[
"Dotti",
"Massimo",
""
],
[
"Contreras",
"Sergio",
""
]
] |
2411.17286 | Spiros Cotsakis | Spiros Cotsakis | Friedmann-Lema\^itre universes and their metamorphoses | v2: 65 pages, more explanations added, published version | Eur. Phys. J. C (2025) 85:579 | 10.1140/epjc/s10052-025-14312-9 | null | gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We analyze the dynamics of the Friedmann-Lema\^itre universes taking into account the different roles played by the fluid parameter and the cosmological constant, as well as the degenerate character of the equations. We find that the Friedmann-Lema\^itre system reduces to four qualitatively inequivalent normal forms and write down the sets of all stable perturbations that may result (the `versal unfoldings'). These sets are of small codimension up to three. We then describe all possible parameter-dependent solutions and their transfigurations to other forms during evolution through the bifurcation sets, these are also fully described. This analysis leads to a picture of cosmological evolution determined by new parameters related to codimension which are zero in standard cosmology. The emerging versal solutions are all free of singularities, while other properties of them are also discussed. | [
{
"version": "v1",
"created": "Tue, 26 Nov 2024 10:17:38 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 11:30:15 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Cotsakis",
"Spiros",
""
]
] |
2411.17888 | Christopher Harvey-Hawes | Christopher Harvey-Hawes and Marco Galoppo | A Novel Test for MOND: Gravitational Lensing by Disc Galaxies | 18 pages, 12 figures, corrected error in phantom dark matter calculation and adjusted results accordingly, updated dark matter and MOND comparison | null | null | null | astro-ph.GA astro-ph.CO gr-qc | http://creativecommons.org/licenses/by/4.0/ | Disc galaxies represent a promising laboratory for the study of gravitational physics, including alternatives to dark matter, owing to the possibility of coupling rotation curves' dynamical data with strong gravitational lensing observations. In particular, Euclid, DES and LSST are predicted to observe hundreds of thousands of gravitational lenses. Here, we investigate disc galaxy strong gravitational lensing in the MOND framework. We employ the concept of equivalent Newtonian systems within the quasi-linear MOND formulation to make use of the standard lensing formalism. We derive the phantom dark matter distribution predicted for realistic disc galaxy models and study the impact of morphological and mass parameters on the expected lensing. We find purely MONDian effects dominate the lensing and generate non-trivial correlations between the lens parameters and the lensing cross section. Moreover, we show that the standard realisation of MOND predicts a substantial increase in the number count of disc galaxy lenses compared to the dark matter-driven predictions, making it distinguishable from the latter in upcoming surveys. Finally, we argue that disc galaxy gravitational lensing, coupled to additional astronomical observations, can be used to constrain the interpolating function of MOND. | [
{
"version": "v1",
"created": "Tue, 26 Nov 2024 21:19:02 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 10:31:57 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Harvey-Hawes",
"Christopher",
""
],
[
"Galoppo",
"Marco",
""
]
] |
2411.18863 | Luke Morgan | Michael Giudici, Luke Morgan and Cheryl E. Praeger | Finite simple groups have many classes of $p$-elements | null | Pacific J. Math. 336 (2025) 137-160 | 10.2140/pjm.2025.336.137 | null | math.GR | http://creativecommons.org/licenses/by/4.0/ | For an element $x$ of a finite group $T$, the $\mathrm{Aut}(T)$-class of $x$ is the set $\{ x^\sigma\mid \sigma\in \mathrm{Aut}(T)\}$. We prove that the order $|T|$ of a finite nonabelian simple group $T$ is bounded above by a function of the parameter $m(T)$, where $m(T)$ is the maximum, over all primes $p$, of the number of $\mathrm{Aut}(T)$-classes of elements of $T$ of $p$-power order. This bound is a substantial generalisation of results of Pyber, and of H\'ethelyi and K\"ulshammer, and it has implications for relative Brauer groups of finite extensions of global fields. | [
{
"version": "v1",
"created": "Thu, 28 Nov 2024 02:06:07 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Giudici",
"Michael",
""
],
[
"Morgan",
"Luke",
""
],
[
"Praeger",
"Cheryl E.",
""
]
] |
2411.19242 | Michael Cummins M.C. | Michael Cummins and Guner Dilsad Er and Michael Muehlebach | Controlling Participation in Federated Learning with Feedback | L4DC 2025 | null | null | null | cs.LG math.OC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We address the problem of client participation in federated learning, where traditional methods typically rely on a random selection of a small subset of clients for each training round. In contrast, we propose FedBack, a deterministic approach that leverages control-theoretic principles to manage client participation in ADMM-based federated learning. FedBack models client participation as a discrete-time dynamical system and employs an integral feedback controller to adjust each client's participation rate individually, based on the client's optimization dynamics. We provide global convergence guarantees for our approach by building on the recent federated learning research. Numerical experiments on federated image classification demonstrate that FedBack achieves up to 50\% improvement in communication and computational efficiency over algorithms that rely on a random selection of clients. | [
{
"version": "v1",
"created": "Thu, 28 Nov 2024 16:26:13 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:32:28 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Cummins",
"Michael",
""
],
[
"Er",
"Guner Dilsad",
""
],
[
"Muehlebach",
"Michael",
""
]
] |
2411.19517 | Tae-Hoon Lee | Tae-Hoon Lee and Min-Soo Kim | RL-SPH: Learning to Achieve Feasible Solutions for Integer Linear Programs | Extended version (20 pages, 7 figures). Accepted at the 2025 AAAI Workshop on AI to Accelerate Science and Engineering (AI2ASE) | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Integer linear programming (ILP) is widely utilized for various combinatorial optimization problems. Primal heuristics play a crucial role in quickly finding feasible solutions for NP-hard ILP. Although \textit{end-to-end learning}-based primal heuristics (E2EPH) have recently been proposed, they are typically unable to independently generate feasible solutions and mainly focus on binary variables. Ensuring feasibility is critical, especially when handling non-binary integer variables. To address this challenge, we propose RL-SPH, a novel reinforcement learning-based start primal heuristic capable of independently generating feasible solutions, even for ILP involving non-binary integers. Experimental results demonstrate that RL-SPH rapidly obtains high-quality feasible solutions, achieving on average a 44x lower primal gap and a 2.3x lower primal integral compared to existing primal heuristics. | [
{
"version": "v1",
"created": "Fri, 29 Nov 2024 07:23:34 GMT"
},
{
"version": "v2",
"created": "Mon, 16 Dec 2024 19:33:38 GMT"
},
{
"version": "v3",
"created": "Tue, 04 Feb 2025 02:43:05 GMT"
},
{
"version": "v4",
"created": "Tue, 11 Mar 2025 08:21:06 GMT"
},
{
"version": "v5",
"created": "Tue, 27 May 2025 08:19:36 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Lee",
"Tae-Hoon",
""
],
[
"Kim",
"Min-Soo",
""
]
] |
2411.19659 | Nikita Belousov | N. Belousov, S. Khoroshkin | Ruijsenaars spectral transform | Expanded Section 6, minor corrections | null | null | null | math-ph hep-th math.CA math.MP nlin.SI | http://creativecommons.org/licenses/by/4.0/ | Spectral decomposition with respect to the wave functions of Ruijsenaars hyperbolic system defines an integral transform, which generalizes classical Fourier integral. For a certain class of analytical symmetric functions we prove inversion formula and orthogonality relations, valid for complex valued parameters of the system. Besides, we study four regimes of unitarity, when this transform defines isomorphisms of the corresponding $L_2$ spaces. | [
{
"version": "v1",
"created": "Fri, 29 Nov 2024 12:32:48 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 06:47:01 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Belousov",
"N.",
""
],
[
"Khoroshkin",
"S.",
""
]
] |
2411.19674 | Guillermo Ballesteros | Guillermo Ballesteros, Jes\'us Gamb\'in Egea, Flavio Riccardi | Finite parts of inflationary loops | Infinitesimal modification on page 12. To appear on JHEP | null | null | IFT UAM-CSIC 24-168 | hep-th astro-ph.CO gr-qc hep-ph | http://creativecommons.org/licenses/by/4.0/ | We present a method for solving loop integrals in dimensional regularization that is particularly useful in the context of inflation. We apply this method to the calculation of the tensor power spectrum induced by scalar fluctuations in slow-roll inflation. | [
{
"version": "v1",
"created": "Fri, 29 Nov 2024 12:58:28 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 21:37:36 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ballesteros",
"Guillermo",
""
],
[
"Egea",
"Jesús Gambín",
""
],
[
"Riccardi",
"Flavio",
""
]
] |
2411.19912 | Samuel Cailleaux | Samuel Cailleaux, Quentin Ficheux, Nicolas Roch, Denis M. Basko | Theory of the photonic Joule effect in superconducting circuits | null | null | 10.1103/b9hw-b6y6 | null | cond-mat.mes-hall cond-mat.supr-con quant-ph | http://creativecommons.org/licenses/by/4.0/ | When a small system is coupled to a bath, it is generally assumed that the state of the bath remains unaffected by the system due to the bath's large number of degrees of freedom. Here we show theoretically that this assumption can be easily violated for photonic baths typically used in experiments involving superconducting circuits. We analyze the dynamics of a voltage-biased Josephson junction coupled to a photonic bath, represented as a long Josephson junction chain. Our findings show that the system can reach a non-equilibrium steady state where the photonic degrees of freedom become significantly overheated, leading to a qualitative change in the current-voltage $I-V$ curve. This phenomenon is analogous to the Joule effect observed in electrical conductors, where flowing current can substantially heat up electrons. Recognizing this effect is crucial for the many applications of high-impedance environments in quantum technologies. | [
{
"version": "v1",
"created": "Fri, 29 Nov 2024 18:17:20 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 09:41:43 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Cailleaux",
"Samuel",
""
],
[
"Ficheux",
"Quentin",
""
],
[
"Roch",
"Nicolas",
""
],
[
"Basko",
"Denis M.",
""
]
] |
2412.00093 | Tim Thyzel | Tim Thyzel | Software-defined lock-in demodulator for low-frequency resistance noise measurements | null | null | null | null | physics.ins-det cond-mat.other | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The resolution of low-frequency resistance noise measurements can be increased by amplitude modulation, shifting the spectrum of the resistance fluctuations away from the 1/f noise contributed by measurement instruments. However, commercial lock-in amplifiers used for de-modulating the fluctuations exhibit a problematic 1/f noise contribution, which imposes a hard lower limit on the relative resistance noise that can be detected. We replace the lock-in amplifier hardware by equivalent digital signal processing performed using open-source software and off-the-shelf data acquisition systems. Compared to previous implementations of the lock-in principle, our solution offers real-time preview capabilities and is resource-efficient for long acquisition times at high sampling rates. Importantly, compared to high-end commercial lock-in instruments, our system offers superior low-frequency noise performance with a reduction of the voltage power spectral density by about two orders of magnitude. | [
{
"version": "v1",
"created": "Wed, 27 Nov 2024 13:29:38 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 10:41:10 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Thyzel",
"Tim",
""
]
] |
2412.00458 | Musfer Adzhymambetov Mr | Musfer Adzhymambetov and Yuri Sinyukov | Extension of the integrated HydroKinetic Model to BES RHIC and GSI-FAIR nuclear collision energies | 11 pages, 10 figures. Version 2 includes an added appendix and minor changes in the main body | null | null | null | hep-ph nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The work presented is devoted to developing the integrated hydrokinetic approach (iHKM) for relativistic nucleus-nucleus collisions. While the previous cycle of works on this topic focused on ultra-relativistic collisions at the top RHIC and different LHC energies, the current work addresses relativistic collisions at lower energies, specifically ranging from approximately 1 to 50 GeV per nucleon pair in the center-of-mass colliding system. The formation times for the initial state of dense matter in such collisions can be up to three orders of magnitude longer than those in ultra-relativistic collisions. This reflects a fundamentally different nature and formation process, particularly concerning the possible stages of initial state evolution, including thermalization (which may be only partial at very low collision energies), subsequent hydrodynamic expansion, and the final transition of matter evolution into a hadronic cascade. These stages, which are fully realized in ultra-relativistic reactions, can also occur within the energy range of BES RHIC, albeit with distinct time scales. This publication not only advances the theoretical development of iHKM (referred to, if necessary, as the extended version of integrated Hydrokinetic Model, iHKMe), but also provides examples of model applications for calculating observables. A systematic description across a wide range of experimental energies, which is preliminary yet quite satisfactory, for spectra, flow, and femtoscopy, will follow this work. | [
{
"version": "v1",
"created": "Sat, 30 Nov 2024 12:16:35 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 09:45:48 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Adzhymambetov",
"Musfer",
""
],
[
"Sinyukov",
"Yuri",
""
]
] |
2412.00551 | Sebastian Seemann | Sebastian Seemann and Francesca Zaffalon | How to stab a polytope | null | Special volume on Positive Geometry, Le Matematiche 80 (1) (2025), 473-493 | null | null | math.CO math.AG | http://creativecommons.org/licenses/by/4.0/ | We study the set of linear subspaces of a fixed dimension intersecting a given polytope. To describe this set as a semialgebraic subset of a Grassmannian, we introduce a Schubert arrangement of the polytope, defined by the Chow forms of the polytope's faces of complementary dimension. We show that the set of subspaces intersecting a specified family of faces is defined by fixing the sign of the Chow forms of their boundaries. We give inequalities defining the set of stabbing subspaces in terms of sign conditions on the Chow form. | [
{
"version": "v1",
"created": "Sat, 30 Nov 2024 18:17:44 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Seemann",
"Sebastian",
""
],
[
"Zaffalon",
"Francesca",
""
]
] |
2412.01028 | Wei Wu | Ze-Zhou Zhang, and Hong-Gang Luo, and Wei Wu | Low-temperature Quantum Metrology Enhanced by Strong Couplings | null | null | 10.1103/PhysRevA.111.052623 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Equilibrium probes have been widely used in various noisy quantum metrology schemes. However, such an equilibrium-probe-based metrology scenario severely suffers from the low-temperature-error divergence problem in the weak-coupling regime. To circumvent this limit, we propose a strategy to eliminate the error-divergence problem by utilizing the strong coupling effects, which can be captured by the reaction-coordinate mapping. The strong couplings induce a noncanonical equilibrium state and greatly enhance the metrology performance. It is found that our metrology precision behaves as a polynomial-type scaling relation, which suggests the reduction of temperature can be used as a resource to improve the metrology performance. Our result is sharply contrary to that of the weakcoupling case, in which the metrology precision exponentially decays as the temperature decreases. Paving a way to realize a high-precision noisy quantum metrology at low temperatures, our result reveals the importance of the non-Markovianity in quantum technologies. | [
{
"version": "v1",
"created": "Mon, 02 Dec 2024 01:22:08 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhang",
"Ze-Zhou",
""
],
[
"Luo",
"Hong-Gang",
""
],
[
"Wu",
"Wei",
""
]
] |
2412.01503 | Xiaoxue Ran | Xiaoxue Ran, Sylvain Capponi, Junchen Rong, Fabien Alet, and Zi Yang Meng | Phase transitions and remnants of fractionalization at finite temperature in the triangular lattice quantum loop model | 8+8 pages, 4+6 figures | null | null | null | cond-mat.str-el cond-mat.stat-mech hep-th quant-ph | http://creativecommons.org/licenses/by/4.0/ | The quantum loop and dimer models are archetypal correlated systems with local constraints. With natural foundations in statistical mechanics, they are of direct relevance to various important physical concepts and systems, such as topological order, lattice gauge theories, geometric frustrations, or more recently Rydberg arrays quantum simulators. However, how the thermal fluctuations interact with constraints has not been explored in the important class of non-bipartite geometries. Here we study, via unbiased quantum Monte Carlo simulations and field theoretical analysis, the finite-temperature phase diagram of the quantum loop model on the triangular lattice. We discover that the recently identified, "hidden" vison plaquette (VP) quantum crystal [1] experiences a finite-temperature continuous transition, which smoothly connects to the (2+1)d Cubic* quantum critical point separating the VP and $\mathbb{Z}_{2}$ quantum spin liquid phases. This finite-temperature phase transition acquires a unique property of ``remnants of fractionalization" at finite temperature, in that, both the cubic order parameter -- the plaquette loop resonance -- and its constituent -- the vison field -- exhibit independent criticality signatures. This phase transition is connected to a 3-state Potts transition between the lattice nematic phase and the high-temperature disordered phase. We discuss the relevance of our results for current experiments on quantum simulation platforms. | [
{
"version": "v1",
"created": "Mon, 02 Dec 2024 13:55:29 GMT"
},
{
"version": "v2",
"created": "Mon, 24 Feb 2025 03:12:48 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 09:25:17 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ran",
"Xiaoxue",
""
],
[
"Capponi",
"Sylvain",
""
],
[
"Rong",
"Junchen",
""
],
[
"Alet",
"Fabien",
""
],
[
"Meng",
"Zi Yang",
""
]
] |
2412.01719 | Gerardo Iuliano | Gerardo Iuliano and Dario Di Nucci | Smart Contract Vulnerabilities, Tools, and Benchmarks: An Updated Systematic Literature Review | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Smart contracts are self-executing programs on blockchain platforms like Ethereum, which have revolutionized decentralized finance by enabling trustless transactions and the operation of decentralized applications. Despite their potential, the security of smart contracts remains a critical concern due to their immutability and transparency, which expose them to malicious actors. Numerous solutions for vulnerability detection have been proposed, but it is still unclear which one is the most effective. This paper presents a systematic literature review that explores vulnerabilities in Ethereum smart contracts, focusing on automated detection tools and benchmark evaluation. We reviewed 3,380 studies from five digital libraries and five major software engineering conferences, applying a structured selection process that resulted in 222 high-quality studies. The key results include a hierarchical taxonomy of 192 vulnerabilities grouped into 14 categories, a comprehensive list of 219 detection tools with corresponding functionalities, methods, and code transformation techniques, a mapping between our taxonomy and the list of tools, and a collection of 133 benchmarks used for tool evaluation. We conclude with a discussion about the insights into the current state of Ethereum smart contract security and directions for future research. | [
{
"version": "v1",
"created": "Mon, 02 Dec 2024 17:08:48 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 20:15:48 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Iuliano",
"Gerardo",
""
],
[
"Di Nucci",
"Dario",
""
]
] |
2412.02059 | Cuihong Wen | Ao Liu, Cuihong Wen, Jieci Wang | Lean classical-quantum hybrid neural network model for image classification | 18 pages,8 figures | Advanced Quantum Technologies. 2025, 2400703 | 10.1002/qute.202400703 | null | quant-ph cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The integration of algorithms from quantum information with neural networks has enabled unprecedented advancements in various domains. Nonetheless, the application of quantum machine learning algorithms for image classification predominantly relies on traditional architectures such as variational quantum circuits. The performance of these models is closely tied to the scale of their parameters, with the substantial demand for parameters potentially leading to limitations in computational resources and a significant increase in computation time. In this paper, we introduce a Lean Classical-Quantum Hybrid Neural Network (LCQHNN), which achieves efficient classification performance with only four layers of variational circuits, thereby substantially reducing computational costs. Our experiments demonstrate that LCQHNN achieves 100\%, 99.02\%, and 85.55\% classification accuracy on MNIST, FashionMNIST, and CIFAR-10 datasets. Under the same parameter conditions, the convergence speed of this method is also faster than that of traditional models. Furthermore, through visualization studies, it is found that the model effectively captures key data features during training and establishes a clear association between these features and their corresponding categories. This study confirms that the employment of quantum algorithms enhances the model's ability to handle complex classification problems. | [
{
"version": "v1",
"created": "Tue, 03 Dec 2024 00:37:11 GMT"
},
{
"version": "v2",
"created": "Mon, 06 Jan 2025 08:38:22 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:23:23 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Liu",
"Ao",
""
],
[
"Wen",
"Cuihong",
""
],
[
"Wang",
"Jieci",
""
]
] |
2412.02255 | Shuai Lu | Shuai Lu, Min Du, P. Victor Debattista | IllustrisTNG Insights: Factors Affecting the Presence of Bars in Disk Galaxies | 19 pages, 16 figures, accepted for publication in A&A | A&A 697, A236 (2025) | 10.1051/0004-6361/202453143 | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bars are important in the secular evolution of galaxies. This study is aimed at exploring the reasons why some galaxies have bars at redshift $z=0$ while others do not. We use ellipse fitting to measure the properties and evolution of bars in the IllustrisTNG cosmological simulation. By using the K-S two-sample test and tracing their evolutionary changes, we analyze the parameter differences between barred and unbarred galaxies. The properties of galaxies with short bars are also studied. When tracing all disk galaxies at $z=0$ back to $z=1$, all of them show similar bar features at $z=1$. The fraction of bars increases in barred and short-bar galaxies but decreases in unbarred galaxies during $z=1-0$. In the case of disk galaxies with stellar mass log$(M_*/M_\odot)> 10.8$, nurture (mainly mergers) plays the most important role in suppressing or destroying bars. Bars are more likely to endure in galaxies that experience fewer mergers, which can be quantified by smaller stellar halos and ex-situ mass fractions. Approximately 60\% of the unbarred galaxies in the local Universe once had a bar. In contrast, the lack of responsiveness to bar instabilities (a larger Toomre-Q parameter) due to a less compact nature plays an important role in generating unbarred disk galaxies with stellar mass log$(M_*/M_\odot)<10.8$. Moreover, short bars generally form at a similar time to normal bars, during which they either grow mildly or contract significantly. The fact that IllustrisTNG simulations produce too many galaxies with short bars indicates that the dynamical properties of the central regions in IllustrisTNG galaxies are less affected by external factors, such as mergers and gas inflows. | [
{
"version": "v1",
"created": "Tue, 03 Dec 2024 08:31:16 GMT"
},
{
"version": "v2",
"created": "Thu, 05 Dec 2024 03:01:15 GMT"
},
{
"version": "v3",
"created": "Sun, 30 Mar 2025 04:10:59 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Lu",
"Shuai",
""
],
[
"Du",
"Min",
""
],
[
"Debattista",
"P. Victor",
""
]
] |
2412.03669 | Murray Moinester | Murray Moinester | Tests of Three-Flavor Chiral Perturbation Theory | 10 pages, 3 tables, Tests of Two-Flavor and Three-Flavor Chiral Perturbation Theory, structured abstract, significantly shorter and more focused discussion, further proofreading and clarification and reference corrections | null | null | null | hep-ph hep-ex | http://creativecommons.org/licenses/by/4.0/ | Introduction: Predictions in the low-energy regime of quantum chromodynamics (QCD) are made using chiral perturbation theory (ChPT). We review ChPT predictions for the polarizabilities of pions and kaons, chiral anomaly (CA) amplitudes $\gamma \rightarrow \pi^-\pi^+ \pi^0$ ($F_{3\pi}$), $\gamma \rightarrow \pi^-\pi^+ \eta$, and $\gamma \rightarrow K K\pi^0$, as well as $\pi^0$ and $\eta$ lifetimes. These quantities are extracted from the small four-momentum transfer Primakoff scattering of high-energy particles (pions, kaons, $\gamma$s) from virtual photons in the Coulomb field of an atomic nucleus. Methods: Ongoing and future Primakoff experiments are needed to obtain these data and to compare them with 2-flavor (u,d) and 3-flavor (u,d,s) ChPT in order to evaluate how well ChPT describes the data. Results: Jefferson Laboratory (JLab) PrimEx $\pi^0$ lifetime, CERN COMPASS $F_{3\pi}$ CA amplitude, and CERN COMPASS pion polarizability data agree with 2-flavor ChPT predictions. This agreement reinforces the identification of the pion as a Goldstone boson. The 3-flavor 2-loop ChPT result for the $\pi^0$ lifetime is approximately 4% lower than the 2-flavor value owing to isospin breaking and mixing of small $\eta$ and $\eta^\prime$ components into the $\pi^0$ wave function. It is 2.4% lower than the PrimEx value, suggesting potential shortcomings in 3-flavor ChPT. Discussion: Transitioning from 2-flavor to 3-flavor ChPT incorporates a strange quark, which is essential for understanding the dynamics of light mesons (pions, kaons, etas). Future CERN AMBER studies of kaon polarizabilities and chiral anomaly amplitudes, as well as JLab studies of the $\eta$ lifetime, are needed to assess the ability of the model to capture the effects of strange quarks. | [
{
"version": "v1",
"created": "Wed, 04 Dec 2024 19:02:26 GMT"
},
{
"version": "v2",
"created": "Thu, 12 Dec 2024 13:15:23 GMT"
},
{
"version": "v3",
"created": "Sun, 22 Dec 2024 12:08:24 GMT"
},
{
"version": "v4",
"created": "Mon, 06 Jan 2025 16:36:43 GMT"
},
{
"version": "v5",
"created": "Wed, 08 Jan 2025 19:14:19 GMT"
},
{
"version": "v6",
"created": "Mon, 20 Jan 2025 09:48:22 GMT"
},
{
"version": "v7",
"created": "Mon, 17 Feb 2025 08:21:06 GMT"
},
{
"version": "v8",
"created": "Tue, 20 May 2025 07:36:47 GMT"
},
{
"version": "v9",
"created": "Tue, 27 May 2025 09:39:26 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Moinester",
"Murray",
""
]
] |
2412.04265 | Yuta Okamoto | Yuta Okamoto and Yuuki Ozaki | On Extrapolation of Treatment Effects in Multiple-Cutoff Regression Discontinuity Designs | null | null | null | null | econ.EM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Regression discontinuity (RD) designs typically identify the treatment effect at a single cutoff point, which may not always be the primary interest. When and how, then, can we learn about treatment effects away from the cutoff? This paper addresses this question within a multiple-cutoff RD framework. We begin by examining the plausibility of the constant bias assumption proposed by Cattaneo, Keele, Titiunik, and Vazquez-Bare (2021) through the lens of rational decision-making behavior. We argue that this assumption is indeed plausible when the groups defined by different cutoffs consist of similar agents and the running variable is effort-invariant, that is, not influenced by the agents' effort (e.g., family income level). However, this positive conclusion no longer holds when the running variable is effort-contingent, meaning it depends on agents' effort (e.g., test scores). In such settings, extrapolated effects based on the constant bias assumption can be biased, even when all other features remain identical except for the cutoff position. To address this issue, we introduce an alternative set of assumptions grounded in empirical motivations and develop a novel partial identification strategy. We illustrate the usefulness and practical relevance of the proposed bounds through empirical examples. In this application, we also find that extrapolation under the constant bias assumption performs reasonably well when the running variable is effort-invariant and some similarity between groups is expected. | [
{
"version": "v1",
"created": "Thu, 05 Dec 2024 15:50:24 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:38:16 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Okamoto",
"Yuta",
""
],
[
"Ozaki",
"Yuuki",
""
]
] |
2412.05219 | Liam Scanlon | Liam J. Scanlon, Santosh Bhusal, Christina M. Hoffmann, Junhong He, Sean R. Parkin, Brennan J. Arnold, William J. Gannon | Structural characterization of the candidate Weyl semimetal CeGaGe | null | Physical Review B. 111 (2025) 184102 | 10.1103/PhysRevB.111.184102 | null | cond-mat.str-el cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Weyl semimetals have a variety of intriguing physical properties, including topologically protected electronic states that coexist with conducting states. Possible exploitation of topologically protected states in a conducting material is promising for technological applications. Weyl semimetals that form in a noncentrosymmetric structure that also contain magnetic moments may host a variety of emergent phenomena that cannot be seen in magnetic, centrosymmetric Weyl materials. It can be difficult to distinguish definitively between a centrosymmetric structure and one of its noncentrosymmetric subgroups with standard powder X-ray diffractometers in cases where two atoms in the compound have nearly the same atomic number, as is the case for the candidate Weyl semimetal CeGaGe. In these cases, a careful single-crystal neutron diffraction experiment with high-angle reflections provides complimentary information to X-ray diffraction and definitively resolves any ambiguity between centrosymmetric and noncentrosymmetric crystal structures. Single-crystal neutron diffraction measurements on the candidate Weyl semimetal CeGaGe confirm that its structure is noncentrosymmetric, described by space group 109 $\left(I4_1md\right)$ rather than the centrosymmetric space group 141 $\left(I4_1/amd\right)$. There are many high-angle reflections in the data set that give clear, physically intuitive evidence that CeGaGe forms with $I4_1md$ symmetry since Bragg planes of these reflections can contain Ga with no Ge or vice versa, whereas the Bragg planes for a structure with $I4_1/amd$ symmetry would have a mix of Ga and Ge. Further, in some crystals we have studied, there is clear evidence for a structural transition from body-centered $I4_1md$ symmetry to primitive $P4_3$ and/or $P4_1$ symmetry. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 17:50:52 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 17:41:29 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Scanlon",
"Liam J.",
""
],
[
"Bhusal",
"Santosh",
""
],
[
"Hoffmann",
"Christina M.",
""
],
[
"He",
"Junhong",
""
],
[
"Parkin",
"Sean R.",
""
],
[
"Arnold",
"Brennan J.",
""
],
[
"Gannon",
"William J.",
""
]
] |
2412.05356 | Beryl Hovis-Afflerbach | B. Hovis-Afflerbach, Y. G\"otberg, A. Schootemeijer, J. Klencki, A. L. Strom, B. A. Ludwig, M. R. Drout | The mass distribution of stars stripped in binaries: The effect of metallicity | 21 pages, 9 figures, (including 2 appendices,) published in A&A | A&A 697, A239 (2025) | 10.1051/0004-6361/202453185 | null | astro-ph.SR astro-ph.GA | http://creativecommons.org/licenses/by/4.0/ | Stars stripped of their hydrogen-rich envelopes through binary interaction are thought to be responsible for both hydrogen-poor supernovae and the hard ionizing radiation observed in low-$Z$ galaxies. A population of these stars was recently observed for the first time, but their prevalence remains unknown. In preparation for such measurements, we estimate the mass distribution of hot, stripped stars using a population synthesis code that interpolates over detailed single and binary stellar evolution tracks. We predict that for a constant star formation rate of $1 \,M_\odot$/yr and regardless of metallicity, a scalable model population contains ~30,000 stripped stars with mass $M_{\rm strip}>1M_\odot$ and ~4,000 stripped stars that are sufficiently massive to explode ($M_{\rm strip}>2.6M_\odot$). Below $M_{\rm strip}=5M_\odot$, the distribution is metallicity-independent and can be described by a power law with the exponent $\alpha \sim -2$. At higher masses and lower metallicity ($Z \lesssim 0.002$), the mass distribution exhibits a drop. This originates from the prediction, frequently seen in evolutionary models, that massive low-metallicity stars do not expand substantially until central helium burning or later and therefore cannot form long-lived stripped stars. With weaker line-driven winds at low metallicity, this suggests that neither binary interaction nor wind mass loss can efficiently strip massive stars at low metallicity. As a result, a "helium-star desert" emerges around $M_{\rm strip} =15\, M_\odot$ at $Z=0.002$, covering an increasingly large mass range with decreasing metallicity. We note that these high-mass stars are those that potentially boost a galaxy's He$^+$-ionizing radiation and that participate in the formation of merging black holes. This "helium-star desert" therefore merits further study. | [
{
"version": "v1",
"created": "Fri, 06 Dec 2024 19:00:00 GMT"
},
{
"version": "v2",
"created": "Thu, 22 May 2025 18:03:42 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Hovis-Afflerbach",
"B.",
""
],
[
"Götberg",
"Y.",
""
],
[
"Schootemeijer",
"A.",
""
],
[
"Klencki",
"J.",
""
],
[
"Strom",
"A. L.",
""
],
[
"Ludwig",
"B. A.",
""
],
[
"Drout",
"M. R.",
""
]
] |
2412.05740 | Lajos Di\'osi | Lajos Di\'osi | Semiclassical world is one of infinite many cloneworlds in common spacetime | 5pp, slightly improved ms, reflection on closely related 1981 paper of Hartle and Horowitz | Phys. Rev. D 110, 084052 (2025) | 10.1103/PhysRevD.111.104065 | null | gr-qc quant-ph | http://creativecommons.org/licenses/by/4.0/ | We consider $N$ clones of the quantized world, interacting with each other via quantum gravity, coupled by the downscaled Newton constant $G/N$. In the limit $N\rightarrow\infty$, we obtain the semiclassical Einstein equation for every single cloneworld. In the non-relativistic limit, De Filippo had already obtained the semiclassical Schr\"odinger-Newton equation, we present an alternative elementary proof. In the general relativistic case we complete the semi-finished derivation of Hartle and Horowitz. We compare our simple correlated cloneworlds with Stamp's more complicated proposal of correlated worldlines and show why they differ despite the conceptual similarity. | [
{
"version": "v1",
"created": "Sat, 07 Dec 2024 20:27:50 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Feb 2025 17:11:32 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Diósi",
"Lajos",
""
]
] |
2412.06372 | Arthur Charlet | A. Charlet, J. Granot, P. Beniamini | Numerical simulations of internal shocks in spherical geometry: hydrodynamics and prompt emission | Accepted for publication in A&A | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | Among the models used to explain the prompt emission of gamma-ray bursts (GRBs), internal shocks is a leading one. Its most basic ingredient is a collision between two cold shells of different Lorentz factors in an ultra-relativistic outflow, which forms a pair of shock fronts that accelerate electrons in their wake. In this model, key features of GRB prompt emission such as the doubly-broken power-law spectral shape arise naturally from the optically-thin synchrotron emission at both shock fronts. We investigate the internal shocks model as a mechanism for prompt emission based on a full hydrodynamical analytic derivation in planar geometry, extending this approach to spherical geometry using hydrodynamic simulations. We used the moving mesh relativistic hydrodynamics code \texttt{GAMMA} to study the collision of two ultra-relativistic cold shells of equal kinetic energy (and power). Using the built-in shock detection, we calculate the corresponding synchrotron emission by the relativistic electrons accelerated into a power-law energy distribution behind the shock, in the fast cooling regime.}During the first dynamical time after the collision, the spherical effects cause the shock strength to decrease with radius. The observed peak frequency decreases faster than expected by other models in the rising part of the pulse, and the peak flux saturates even for moderately short pulses. This is likely caused by the very sharp edges of the shells in our model, while smoother edges will probably mitigate this effect. Our model traces the evolution of the peak frequency back to the source activity time scales. | [
{
"version": "v1",
"created": "Mon, 09 Dec 2024 10:50:35 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 08:33:23 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Charlet",
"A.",
""
],
[
"Granot",
"J.",
""
],
[
"Beniamini",
"P.",
""
]
] |
2412.06620 | Iris van Gemeren | Gast\'on Creci, Iris van Gemeren, Tanja Hinderer, Jan Steinhoff | Tidal effects in gravitational waves from neutron stars in scalar-tensor theories of gravity | v2:revised manuscript for resubmission, minor corrections, 22 pages + appendix, 11 figures | SciPost Phys. Core 8, 042 (2025) | 10.21468/SciPostPhysCore.8.2.042 | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We compute tidal signatures in the gravitational waves (GWs) from neutron star binary inspirals in scalar-tensor gravity, where the dominant adiabatic even-parity tidal interactions involve three types of Love numbers that depend on the matter equation of state and parameters of the gravitational theory. We calculate the modes of the GW amplitudes and the phase evolution in the time and frequency domain, working up to first order in the post-Newtonian and small finite-size approximations. We also perform several case studies to quantify the dipolar and quadrupolar tidal effects and their parameter dependencies specialized to Gaussian couplings. We show that various tidal contributions enter with different signs and scalings with frequency, which generally leads to smaller net tidal GW imprints than for the same binary system in General Relativity. | [
{
"version": "v1",
"created": "Mon, 09 Dec 2024 16:14:08 GMT"
},
{
"version": "v2",
"created": "Mon, 14 Apr 2025 08:07:43 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Creci",
"Gastón",
""
],
[
"van Gemeren",
"Iris",
""
],
[
"Hinderer",
"Tanja",
""
],
[
"Steinhoff",
"Jan",
""
]
] |
2412.07250 | Santiago Francisco Caballero-Benitez | Brahyam R\'ios-S\'anchez and Santiago F. Caballero-Ben\'itez | Control, competition and coexistence of effective magnetic orders by interactions in Bose-Einstein condensates with high-Q cavities | 29 pages, 14 figures | null | null | null | cond-mat.quant-gas physics.atom-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ultracold atomic systems confined in optical cavities have been demonstrated as a laboratory for the control of quantum matter properties and analog quantum simulation. Often neglected, but soon amenable to manipulation in a new generation of experiments, we show that atomic many-body interactions allow additional control in the cavity driven self-organization of effective spinor Bose-Einstein condensates (BEC). We theoretically show that a rich landscape of magnetic ordering configurations emerges. This can be controlled by modifying the geometry of the light-fields in the system with the interplay of two-body interactions and the cavity induced interactions. This leads to competition scenarios and phase separated dynamics. Our results show that it is possible to tailor on demand configurations possibly useful for analog quantum simulation of magnetic materials with highly controllable parameters in a single experimentally realistic setup. | [
{
"version": "v1",
"created": "Tue, 10 Dec 2024 07:20:26 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 22:23:59 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ríos-Sánchez",
"Brahyam",
""
],
[
"Caballero-Benítez",
"Santiago F.",
""
]
] |
2412.07632 | Sebastian Bor\'owka | Bartosz Kasza, Sebastian Bor\'owka, Wojciech Wasilewski, Micha{\l} Parniak | Atomic-optical interferometry in fractured loops: a general solution for Rydberg radio frequency receivers | 11 pages, 12 figures | Phys. Rev. A 111, 053718 (2025) | 10.1103/PhysRevA.111.053718 | null | physics.atom-ph physics.optics quant-ph | http://creativecommons.org/licenses/by/4.0/ | The development of novel radio frequency atomic receivers brings attention to the theoretical description of atom-light interactions in sophisticated, multilevel schemes. Of special interest, are the schemes where several interaction paths interfere with each other, bringing about the phase-sensitive measurement of detected radio fields. In the theoretical modeling of those cases, the common assumptions are often insufficient to determine the boundary detection parameters, such as receiving bandwidth or saturation point, critical for practical considerations of atomic sensing technology. This evokes the resurfacing of a long-standing problem on how to describe an atom-light interaction in a fractured loop. In such a case, the quantum steady state is not achieved even with constant, continuous interactions. Here we propose a method for modeling of such a system, basing our approach on the Fourier expansion of a non-equilibrium steady state. The proposed solution is both numerically effective and able to predict edge cases, such as saturation. Furthermore, as an example, we employ this method to provide a complete description of a Rydberg superheterodyne receiver, obtaining the boundary parameters describing the operation of this atomic detector. | [
{
"version": "v1",
"created": "Tue, 10 Dec 2024 16:13:48 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Dec 2024 13:12:18 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kasza",
"Bartosz",
""
],
[
"Borówka",
"Sebastian",
""
],
[
"Wasilewski",
"Wojciech",
""
],
[
"Parniak",
"Michał",
""
]
] |
2412.07651 | Hannes Mildner | Hannes Mildner | An EWPD SMEFT likelihood for the LHC -- and how to improve it with measurements of W and Z boson properties | 43 pages, 12 figures, 8 tables; v2: added Figure to illustrate most important EWPOs per operator, added comparison to SMEFiT, extended discussion of PDF uncertainties, fixed typos, improved colour scheme | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a computer code for analyzing electroweak precision data (EWPD) in the framework of the Standard Model Effective Field Theory (SMEFT), highlights the importance of recent ATLAS and CMS precision measurements, and introduces a novel analysis of the forward-backward asymmetry at the LHC. The computer code provides the likelihood of SMEFT Wilson coefficients based on precision measurements of $W$ and $Z$ pole observables, interpolation formulas for Standard Model predictions, and modular SMEFT parametrizations. SMEFT predictions including next-to-leading-order (NLO) effects in perturbative and SMEFT expansion are available and five alternative electroweak input parameter schemes are supported. The likelihood addresses shortcomings of previous formulations in the treatment of parametric uncertainties and can be straightforwardly included in SMEFT fits of LHC data. The input parameter scheme dependence and role of NLO corrections is studied for the EWPD fit in the SMEFT. Furthermore, the impact of recent ATLAS and CMS measurements - of the $W$ boson mass and width, of the lepton flavour universality (LFU) of $W$ branching fractions, and the effective leptonic weak mixing angle - is analyzed. A test of LFU that surpasses the precision of existing measurements is proposed based on the weak mixing angle measurement. Finally, an ATLAS Drell-Yan triple-differential cross-section measurement is reinterpreted in the SMEFT and combined with the EWPD likelihood. This analysis demonstrates the feasibility of the LFU precision test, improves constraints on muon couplings with respect to the world average, and determines a combination of the quark-coupling asymmetry parameters $A_u$ and $A_d$ with a precision comparable to that of the heavy flavour parameters $A_c$ and $A_b$. | [
{
"version": "v1",
"created": "Tue, 10 Dec 2024 16:37:26 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:11:41 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Mildner",
"Hannes",
""
]
] |
2412.08106 | Ian Weng Chan | Ian Weng Chan, Andrei L. Badescu, X. Sheldon Lin | Assessing Driving Risk Through Unsupervised Detection of Anomalies in Telematics Time Series Data | null | ASTIN Bull. 55 (2025) 205-241 | 10.1017/asb.2025.14 | null | stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vehicle telematics provides granular data for dynamic driving risk assessment, but current methods often rely on aggregated metrics (e.g., harsh braking counts) and do not fully exploit the rich time-series structure of telematics data. In this paper, we introduce a flexible framework using continuous-time hidden Markov model (CTHMM) to model and analyze trip-level telematics data. Unlike existing methods, the CTHMM models raw time-series data without predefined thresholds on harsh driving events or assumptions about accident probabilities. Moreover, our analysis is based solely on telematics data, requiring no traditional covariates such as driver or vehicle characteristics. Through unsupervised anomaly detection based on pseudo-residuals, we identify deviations from normal driving patterns -- defined as the prevalent behaviour observed in a driver's history or across the population -- which are linked to accident risk. Validated on both controlled and real-world datasets, the CTHMM effectively detects abnormal driving behaviour and trips with increased accident likelihood. In real data analysis, higher anomaly levels in longitudinal and lateral accelerations consistently correlate with greater accident risk, with classification models using this information achieving ROC-AUC values as high as 0.86 for trip-level analysis and 0.78 for distinguishing drivers with claims. Furthermore, the methodology reveals significant behavioural differences between drivers with and without claims, offering valuable insights for insurance applications, accident analysis, and prevention. | [
{
"version": "v1",
"created": "Wed, 11 Dec 2024 05:12:25 GMT"
},
{
"version": "v2",
"created": "Mon, 31 Mar 2025 18:10:58 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Chan",
"Ian Weng",
""
],
[
"Badescu",
"Andrei L.",
""
],
[
"Lin",
"X. Sheldon",
""
]
] |
2412.09342 | Ralf R\"omer | Ralf R\"omer, Alexander von Rohr, Angela P. Schoellig | Diffusion Predictive Control with Constraints | Accepted to L4DC 2025. Code: https://github.com/ralfroemer99/dpcc. 14 pages, 3 figures, 3 tables | null | null | null | cs.RO cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion models have become popular for policy learning in robotics due to their ability to capture high-dimensional and multimodal distributions. However, diffusion policies are stochastic and typically trained offline, limiting their ability to handle unseen and dynamic conditions where novel constraints not represented in the training data must be satisfied. To overcome this limitation, we propose diffusion predictive control with constraints (DPCC), an algorithm for diffusion-based control with explicit state and action constraints that can deviate from those in the training data. DPCC incorporates model-based projections into the denoising process of a trained trajectory diffusion model and uses constraint tightening to account for model mismatch. This allows us to generate constraint-satisfying, dynamically feasible, and goal-reaching trajectories for predictive control. We show through simulations of a robot manipulator that DPCC outperforms existing methods in satisfying novel test-time constraints while maintaining performance on the learned control task. | [
{
"version": "v1",
"created": "Thu, 12 Dec 2024 15:10:22 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:45:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Römer",
"Ralf",
""
],
[
"von Rohr",
"Alexander",
""
],
[
"Schoellig",
"Angela P.",
""
]
] |
2412.10312 | Federico Ruggeri | Federico Ruggeri, Gaetano Signorelli | Interlocking-free Selective Rationalization Through Genetic-based Learning | null | null | null | null | cs.LG cs.AI cs.CL cs.NE | http://creativecommons.org/licenses/by-sa/4.0/ | A popular end-to-end architecture for selective rationalization is the select-then-predict pipeline, comprising a generator to extract highlights fed to a predictor. Such a cooperative system suffers from suboptimal equilibrium minima due to the dominance of one of the two modules, a phenomenon known as interlocking. While several contributions aimed at addressing interlocking, they only mitigate its effect, often by introducing feature-based heuristics, sampling, and ad-hoc regularizations. We present GenSPP, the first interlocking-free architecture for selective rationalization that does not require any learning overhead, as the above-mentioned. GenSPP avoids interlocking by performing disjoint training of the generator and predictor via genetic global search. Experiments on a synthetic and a real-world benchmark show that our model outperforms several state-of-the-art competitors. | [
{
"version": "v1",
"created": "Fri, 13 Dec 2024 17:52:48 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 15:07:23 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Ruggeri",
"Federico",
""
],
[
"Signorelli",
"Gaetano",
""
]
] |
2412.10526 | Janet Zhong Miss | Janet Zhong, Heming Wang, Alexander N Poddubny, Shanhui Fan | Topological nature of edge states for one-dimensional systems without symmetry protection | comments welcome :) | null | null | null | cond-mat.mes-hall quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We numerically verify and analytically prove a winding number invariant that correctly predicts the number of edge states in one-dimensional, nearest-neighbor (between unit cells), two-band models with any complex couplings and open boundaries. Our winding number uses analytical continuation of the wave-vector into the complex plane and involves two special points on the full Riemann surface band structure that correspond to bulk eigenvector degeneracies. Our winding number is invariant under unitary or similarity transforms. We emphasize that the topological criteria we propose here differ from what is traditionally defined as a topological or trivial phase in symmetry-protected classification studies. It is a broader invariant for our model that supports nonzero energy edge states and its transition does not coincide with the gap closing condition. When the relevant symmetries are applied, our invariant reduces to well-known Hermitian and non-Hermitian symmetry-protected topological invariants. | [
{
"version": "v1",
"created": "Fri, 13 Dec 2024 19:44:54 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 23:31:06 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Zhong",
"Janet",
""
],
[
"Wang",
"Heming",
""
],
[
"Poddubny",
"Alexander N",
""
],
[
"Fan",
"Shanhui",
""
]
] |
2412.10792 | Sertac Kilickaya | Sertac Kilickaya, Mete Ahishali, Cansu Celebioglu, Fahad Sohrab, Levent Eren, Turker Ince, Murat Askar, Moncef Gabbouj | Audio-based Anomaly Detection in Industrial Machines Using Deep One-Class Support Vector Data Description | To be published in 2025 IEEE Symposium Series on Computational Intelligence | null | 10.1109/CIESCompanion65073.2025.11010815 | null | cs.SD cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The frequent breakdowns and malfunctions of industrial equipment have driven increasing interest in utilizing cost-effective and easy-to-deploy sensors, such as microphones, for effective condition monitoring of machinery. Microphones offer a low-cost alternative to widely used condition monitoring sensors with their high bandwidth and capability to detect subtle anomalies that other sensors might have less sensitivity. In this study, we investigate malfunctioning industrial machines to evaluate and compare anomaly detection performance across different machine types and fault conditions. Log-Mel spectrograms of machinery sound are used as input, and the performance is evaluated using the area under the curve (AUC) score for two different methods: baseline dense autoencoder (AE) and one-class deep Support Vector Data Description (deep SVDD) with different subspace dimensions. Our results over the MIMII sound dataset demonstrate that the deep SVDD method with a subspace dimension of 2 provides superior anomaly detection performance, achieving average AUC scores of 0.84, 0.80, and 0.69 for 6 dB, 0 dB, and -6 dB signal-to-noise ratios (SNRs), respectively, compared to 0.82, 0.72, and 0.64 for the baseline model. Moreover, deep SVDD requires 7.4 times fewer trainable parameters than the baseline dense AE, emphasizing its advantage in both effectiveness and computational efficiency. | [
{
"version": "v1",
"created": "Sat, 14 Dec 2024 11:05:06 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kilickaya",
"Sertac",
""
],
[
"Ahishali",
"Mete",
""
],
[
"Celebioglu",
"Cansu",
""
],
[
"Sohrab",
"Fahad",
""
],
[
"Eren",
"Levent",
""
],
[
"Ince",
"Turker",
""
],
[
"Askar",
"Murat",
""
],
[
"Gabbouj",
"Moncef",
""
]
] |
2412.11189 | Bugeun Kim | Byungjun Kim, Minju Kim, Dayeon Seo, Bugeun Kim | Leveraging Large Language Models for Active Merchant Non-player Characters | Accepted to IJCAI 2025 | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | We highlight two significant issues leading to the passivity of current merchant non-player characters (NPCs): pricing and communication. While immersive interactions with active NPCs have been a focus, price negotiations between merchant NPCs and players remain underexplored. First, passive pricing refers to the limited ability of merchants to modify predefined item prices. Second, passive communication means that merchants can only interact with players in a scripted manner. To tackle these issues and create an active merchant NPC, we propose a merchant framework based on large language models (LLMs), called MART, which consists of an appraiser module and a negotiator module. We conducted two experiments to explore various implementation options under different training methods and LLM sizes, considering a range of possible game environments. Our findings indicate that finetuning methods, such as supervised finetuning (SFT) and knowledge distillation (KD), are effective in using smaller LLMs to implement active merchant NPCs. Additionally, we found three irregular cases arising from the responses of LLMs. | [
{
"version": "v1",
"created": "Sun, 15 Dec 2024 13:48:39 GMT"
},
{
"version": "v2",
"created": "Wed, 08 Jan 2025 11:24:17 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 16:23:46 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Kim",
"Byungjun",
""
],
[
"Kim",
"Minju",
""
],
[
"Seo",
"Dayeon",
""
],
[
"Kim",
"Bugeun",
""
]
] |
2412.11257 | Fengpei Li | Fengpei Li, Haoxian Chen, Jiahe Lin, Arkin Gupta, Xiaowei Tan, Honglei Zhao, Gang Xu, Yuriy Nevmyvaka, Agostino Capponi, Henry Lam | Prediction-Enhanced Monte Carlo: A Machine Learning View on Control Variate | null | null | null | null | stat.ML cs.CE cs.LG q-fin.PR | http://creativecommons.org/licenses/by/4.0/ | For many complex simulation tasks spanning areas such as healthcare, engineering, and finance, Monte Carlo (MC) methods are invaluable due to their unbiased estimates and precise error quantification. Nevertheless, Monte Carlo simulations often become computationally prohibitive, especially for nested, multi-level, or path-dependent evaluations lacking effective variance reduction techniques. While machine learning (ML) surrogates appear as natural alternatives, naive replacements typically introduce unquantifiable biases. We address this challenge by introducing Prediction-Enhanced Monte Carlo (PEMC), a framework that leverages modern ML models as learned predictors, using cheap and parallelizable simulation as features, to output unbiased evaluation with reduced variance and runtime. PEMC can also be viewed as a "modernized" view of control variates, where we consider the overall computation-cost-aware variance reduction instead of per-replication reduction, while bypassing the closed-form mean function requirement and maintaining the advantageous unbiasedness and uncertainty quantifiability of Monte Carlo.
We illustrate PEMC's broader efficacy and versatility through three examples: first, equity derivatives such as variance swaps under stochastic local volatility models; second, interest rate derivatives such as swaption pricing under the Heath-Jarrow-Morton (HJM) interest-rate model. Finally, we showcase PEMC in a socially significant context - ambulance dispatch and hospital load balancing - where accurate mortality rate estimates are key for ethically sensitive decision-making. Across these diverse scenarios, PEMC consistently reduces variance while preserving unbiasedness, highlighting its potential as a powerful enhancement to standard Monte Carlo baselines. | [
{
"version": "v1",
"created": "Sun, 15 Dec 2024 17:41:38 GMT"
},
{
"version": "v2",
"created": "Mon, 26 May 2025 21:29:46 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Li",
"Fengpei",
""
],
[
"Chen",
"Haoxian",
""
],
[
"Lin",
"Jiahe",
""
],
[
"Gupta",
"Arkin",
""
],
[
"Tan",
"Xiaowei",
""
],
[
"Zhao",
"Honglei",
""
],
[
"Xu",
"Gang",
""
],
[
"Nevmyvaka",
"Yuriy",
""
],
[
"Capponi",
"Agostino",
""
],
[
"Lam",
"Henry",
""
]
] |
2412.11283 | Felix Lotter | Felix Lotter, Rosa Prei{\ss} | Cyclic polytopes through the lens of iterated integrals | 20 pages, comments are welcome! | Special volume on Positive Geometry, Le Matematiche 80 (1) (2025), 365-385 | null | null | math.RA math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The volume of a cyclic polytope can be obtained by forming an iterated integral, known as the path signature, along a suitable piecewise linear path running through its edges. Different choices of such a path are related by the action of a subgroup of the combinatorial automorphisms of the polytope. Motivated by this observation, we look for other polynomials in the vertices of a cyclic polytope that arise as path signatures and are invariant under the subgroup action. We prove that there are infinitely many such invariants which are algebraically independent in the shuffle algebra. | [
{
"version": "v1",
"created": "Sun, 15 Dec 2024 19:08:29 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Lotter",
"Felix",
""
],
[
"Preiß",
"Rosa",
""
]
] |
2412.11429 | Yi Gao | Yi Gao | Theoretical investigation of the superconducting pairing symmetry in a bilayer two-orbital model of pressurized La$_3$Ni$_2$O$_7$ | null | J. Phys.: Condens. Matter 37, 235601 (2025) | 10.1088/1361-648X/add6fc | null | cond-mat.supr-con | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the superconducting pairing symmetry in pressurized La$_3$Ni$_2$O$_7$ based on a bilayer two-orbital model. There are two symmetric bands $\alpha$ and $\gamma$, as well as two antisymmetric ones $\beta$ and $\delta$. It is found that the $\gamma$ band induces considerable ferromagnetic spin fluctuation and prefers an odd-frequency, $s$-wave spin triplet pairing state. The addition of the other bands gradually suppresses the ferromagnetic spin fluctuation and enhances the antiferromagnetic ones. The superconducting pairing then evolves from a spin triplet into a $d$-wave spin singlet, and finally into an $s_\pm$-wave one. The competition between the $d$-wave and $s_\pm$-wave pairings relies on whether the ferromagnetic spin fluctuation is suppressed completely or not. | [
{
"version": "v1",
"created": "Mon, 16 Dec 2024 03:57:39 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Gao",
"Yi",
""
]
] |
2412.11569 | Michele Ceriotti | Filippo Bigi, Marcel Langer and Michele Ceriotti | The dark side of the forces: assessing non-conservative force models for atomistic machine learning | 10 pages (including references) + appendix Conference format | null | null | null | physics.chem-ph cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The use of machine learning to estimate the energy of a group of atoms, and the forces that drive them to more stable configurations, has revolutionized the fields of computational chemistry and materials discovery. In this domain, rigorous enforcement of symmetry and conservation laws has traditionally been considered essential. For this reason, interatomic forces are usually computed as the derivatives of the potential energy, ensuring energy conservation. Several recent works have questioned this physically constrained approach, suggesting that directly predicting the forces yields a better trade-off between accuracy and computational efficiency -- and that energy conservation can be learned during training. This work investigates the applicability of such non-conservative models in microscopic simulations. We identify and demonstrate several fundamental issues, from ill-defined convergence of geometry optimization to instability in various types of molecular dynamics. Contrary to the case of rotational symmetry, energy conservation is hard to learn, monitor, and correct for. The best approach to exploit the acceleration afforded by direct force prediction might be to use it in tandem with a conservative model, reducing -- rather than eliminating -- the additional cost of backpropagation, but avoiding the pathological behavior associated with non-conservative forces. | [
{
"version": "v1",
"created": "Mon, 16 Dec 2024 08:55:03 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Feb 2025 19:48:32 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 14:09:45 GMT"
}
] | 2025-05-28T00:00:00 | [
[
"Bigi",
"Filippo",
""
],
[
"Langer",
"Marcel",
""
],
[
"Ceriotti",
"Michele",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.