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
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2509.03758
|
Alvaro Almeida Gomez
|
Alvaro Almeida Gomez
|
Learning functions through Diffusion Maps
|
Comments are welcome
| null | null | null |
cs.LG cs.NA math.NA
|
http://creativecommons.org/licenses/by/4.0/
|
We propose a data-driven method for approximating real-valued functions on smooth manifolds, building on the Diffusion Maps framework under the manifold hypothesis. Given pointwise evaluations of a function, the method constructs a smooth extension to the ambient space by exploiting diffusion geometry and its connection to the heat equation and the Laplace-Beltrami operator.
To address the computational challenges of high-dimensional data, we introduce a dimensionality reduction strategy based on the low-rank structure of the distance matrix, revealed via singular value decomposition (SVD). In addition, we develop an online updating mechanism that enables efficient incorporation of new data, thereby improving scalability and reducing computational cost.
Numerical experiments, including applications to sparse CT reconstruction, demonstrate that the proposed methodology outperforms classical feedforward neural networks and interpolation methods in terms of both accuracy and efficiency.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 22:57:33 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Gomez",
"Alvaro Almeida",
""
]
] |
2509.03759
|
Rufus Willett
|
Robin Deeley and Rufus Willett
|
A trace pairing and Elliott invariant for groupoid homology
| null | null | null | null |
math.OA math.AT math.DS math.KT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
For an \'{e}tale groupoid, we define a pairing between the Crainic-Moerdijk groupoid homology and the simplex of invariant Borel probability measures on the base space. The main novelty here is that the groupoid need not have totally disconnected base space, and thus the pairing can give more refined information than the measures of clopen subsets of the base space.
Our principal motivation is $C^*$-algebra theory. The Elliott invariant of a $C^*$-algebra is defined in terms of $K$-theory and traces; it is fundamental in the long-running program to classify simple $C^*$-algebras (satisfying additional necessary conditions). We use our pairing to define a groupoid Elliott invariant, and show that for many interesting groupoids it agrees with the $C^*$-algebraic Elliott invariant of the groupoid $C^*$-algebra: this includes irrational rotation algebras and the $C^*$-algebras arising from orbit breaking constructions studied by the first listed author, Putnam, and Strung. These results can be thought of as establishing a refinement of Matui's HK conjecture for the relevant groupoids.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 22:58:31 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Deeley",
"Robin",
""
],
[
"Willett",
"Rufus",
""
]
] |
2509.03760
|
Ariel A. P\'erez
|
Rodrigo Lecaros, Ariel A. P\'erez, Manuel F. Prado
|
Inverse problem for semidiscrete stochastic parabolic operators in arbitrary dimensions
| null | null | null | null |
math.AP
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we study two types of inverse problems for space semidiscrete stochastic parabolic equations in arbitrary dimensions. The first problem concerns a semidiscrete inverse source problem, which involves determining the random source term of the white noise in the stochastic parabolic equation using observation data of the solution at the terminal time and the trace of the discrete spatial derivative of the solution on a subdomain at the boundary of the space. The second problem addresses a semidiscrete Cauchy inverse problem, which involves determining the solution of the stochastic parabolic equation in a subdomain of space-time, from the data of the solution and the trace of the discrete spatial derivative on a subdomain of the boundary of space and time. To address these problems, we establish two Carleman estimates for a semidiscrete stochastic parabolic operator in arbitrary dimensions: one for the homogeneous boundary condition and the other for the nonhomogeneous boundary condition. Applying these Carleman estimates, we obtain Lipschitz and H\"older stability for the first and second inverse problems, respectively.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:01:53 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Lecaros",
"Rodrigo",
""
],
[
"Pérez",
"Ariel A.",
""
],
[
"Prado",
"Manuel F.",
""
]
] |
2509.03761
|
Joseph Rich
|
Joseph Rich, Conrad Oakes, Lior Pachter
|
Optimizing alluvial plots
|
18 pages, 3 main figures, 6 supplementary figures, 1 supplementary table
| null | null | null |
cs.CG
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Alluvial plots can be effective for visualization of multivariate data, but rely on ordering of alluvia that can be non-trivial to arrange. We formulate two optimization problems that formalize the challenge of ordering and coloring partitions in alluvial plots. While solving these optimization problems is challenging in general, we show that the NeighborNet algorithm from phylogenetics can be adapted to provide excellent results in typical use cases. Our methods are implemented in a freely available R package available on GitHub at https://github.com/pachterlab/wompwomp
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:02:34 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Rich",
"Joseph",
""
],
[
"Oakes",
"Conrad",
""
],
[
"Pachter",
"Lior",
""
]
] |
2509.03762
|
Sathwik P Chadaga
|
Sathwik Chadaga and Eytan Modiano
|
Drift Plus Optimistic Penalty -- A Learning Framework for Stochastic Network Optimization
| null | null | null | null |
cs.NI cs.SY eess.SY
|
http://creativecommons.org/licenses/by-sa/4.0/
|
We consider the problem of joint routing and scheduling in queueing networks, where the edge transmission costs are unknown. At each time-slot, the network controller receives noisy observations of transmission costs only for those edges it selects for transmission. The network controller's objective is to make routing and scheduling decisions so that the total expected cost is minimized. This problem exhibits an exploration-exploitation trade-off, however, previous bandit-style solutions cannot be directly applied to this problem due to the queueing dynamics. In order to ensure network stability, the network controller needs to optimize throughput and cost simultaneously. We show that the best achievable cost is lower bounded by the solution to a static optimization problem, and develop a network control policy using techniques from Lyapunov drift-plus-penalty optimization and multi-arm bandits. We show that the policy achieves a sub-linear regret of order $O(\sqrt{T}\log T)$, as compared to the best policy that has complete knowledge of arrivals and costs. Finally, we evaluate the proposed policy using simulations and show that its regret is indeed sub-linear.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:06:43 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chadaga",
"Sathwik",
""
],
[
"Modiano",
"Eytan",
""
]
] |
2509.03763
|
Michela Negro
|
Michela Negro, Zorawar Wadiasingh, George Younes, Eric Burns, Anirudh Patel, Brian D. Metzger, Todd A. Thompson, Daryl Haggard, S. Bradley Cenko
|
Fast X-ray Transient detection with AXIS: application to Magnetar Giant Flares
|
Submitted to Open Journal of Astrophysics
| null | null | null |
astro-ph.HE
|
http://creativecommons.org/licenses/by/4.0/
|
Magnetar giant flares (MGFs) are among the most luminous high-energy transients in the local universe, consisting of a short, intense MeV gamma-ray spike followed by a softer, pulsating X-ray tail and possibly delayed radioactive emission. While only three Galactic events have been firmly detected, several extragalactic candidates have recently been reported, motivating the need for sensitive, rapid-response gamma- and X-ray facilities to constrain their rates and energetics. We present a feasibility study of detecting MGFs with the Advanced X-ray Imaging Satellite (AXIS), focusing on two complementary pathways: (i) serendipitous discovery of the prompt gamma-ray spike within the field of view, and (ii) rapid follow-up of MGF tails in nearby galaxies. Using sensitivity rescaling and volumetric rate estimates, we find that serendipitous detection of prompt spikes during the mission lifetime is possible but unlikely, given their hardness and short duration. In contrast, AXIS's superior sensitivity, if accompanied by fast repointing capabilities, offer an extraordinary opportunity to detect pulsating X-ray tails out to about 20 Mpc, enabling the first extragalactic measurements of periodic modulations and potentially constraining emission geometry and fireball physics. Finally, we evaluate the detectability of soft X-ray line emission from r-process nucleosynthesis in MGFs, finding that such signals are extremely faint and confining the detection to Galactic distances. Our study offer a general framework for assessing the detectability of short transients with future missions.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:07:34 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Negro",
"Michela",
""
],
[
"Wadiasingh",
"Zorawar",
""
],
[
"Younes",
"George",
""
],
[
"Burns",
"Eric",
""
],
[
"Patel",
"Anirudh",
""
],
[
"Metzger",
"Brian D.",
""
],
[
"Thompson",
"Todd A.",
""
],
[
"Haggard",
"Daryl",
""
],
[
"Cenko",
"S. Bradley",
""
]
] |
2509.03764
|
Han Wang
|
Han Wang, Alex Whitworth, Pak Ming Cheung, Zhenjie Zhang, Krishna Kamath
|
LLM-based Relevance Assessment for Web-Scale Search Evaluation at Pinterest
|
RecSys 2025 EARL Workshop
| null | null | null |
cs.IR cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Relevance evaluation plays a crucial role in personalized search systems to ensure that search results align with a user's queries and intent. While human annotation is the traditional method for relevance evaluation, its high cost and long turnaround time limit its scalability. In this work, we present our approach at Pinterest Search to automate relevance evaluation for online experiments using fine-tuned LLMs. We rigorously validate the alignment between LLM-generated judgments and human annotations, demonstrating that LLMs can provide reliable relevance measurement for experiments while greatly improving the evaluation efficiency. Leveraging LLM-based labeling further unlocks the opportunities to expand the query set, optimize sampling design, and efficiently assess a wider range of search experiences at scale. This approach leads to higher-quality relevance metrics and significantly reduces the Minimum Detectable Effect (MDE) in online experiment measurements.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:07:49 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Wang",
"Han",
""
],
[
"Whitworth",
"Alex",
""
],
[
"Cheung",
"Pak Ming",
""
],
[
"Zhang",
"Zhenjie",
""
],
[
"Kamath",
"Krishna",
""
]
] |
2509.03765
|
Ilya Nemenman
|
Ilya Nemenman and Pankaj Mehta
|
Randomness with constraints: constructing minimal models for high-dimensional biology
|
9 pages, 2 figures
| null | null | null |
physics.bio-ph q-bio.QM
|
http://creativecommons.org/licenses/by/4.0/
|
Biologists and physicists have a rich tradition of modeling living systems with simple models composed of a few interacting components. Despite the remarkable success of this approach, it remains unclear how to use such finely tuned models to study complex biological systems composed of numerous heterogeneous, interacting components. One possible strategy for taming this biological complexity is to embrace the idea that many biological behaviors we observe are ``typical'' and can be modeled using random systems that respect biologically-motivated constraints. Here, we review recent works showing how this approach can be used to make close connection with experiments in biological systems ranging from neuroscience to ecology and evolution and beyond. Collectively, these works suggest that the ``random-with-constraints'' paradigm represents a promising new modeling strategy for capturing experimentally observed dynamical and statistical features in high-dimensional biological data and provides a powerful minimal modeling philosophy for biology.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:13:02 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Nemenman",
"Ilya",
""
],
[
"Mehta",
"Pankaj",
""
]
] |
2509.03766
|
Achraf Khoudiri
|
Achraf Khoudiri, Abderrahman Oularabi, Khadija El Anouz, \.Ilkay Demir, Abderrahim El Allati
|
Coherence-Driven Quantum Battery Charging via Autonomous Thermal Machines: Energy Transfer, Memory Effects, and Ergotropy Enhancement
|
This work has been submitted as a preprint on arXiv and is intended for future publication in a peer-reviewed journal. The authors retain all rights for formal publication
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
In this work, we study a hybrid quantum system composed of a quantum battery and a coherence-driven charger interacting with a Quantum Autonomous Thermal Machine (QATM). The QATM, made of two qubits each coupled to Markovian Bosonics thermal reservoirs at different temperatures, acts as a structured environment that mediates energy and coherence transfer between the charger and the battery. By presenting a coherent driving field on the charger, we investigate the coherence injection effect on the dynamics, including non-Markovianity, power of charging, coherence storage, and ergotropy. We show that the QATM effectively filters the decoherence induced by the thermal baths and induces non-Markovian memory effects due to correlation backflow. Our results demonstrate that coherence driving enhances the battery's ergotropy and power of charging by preserving the internal energy of the charger and facilitating coherent energy transfer.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:16:26 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Khoudiri",
"Achraf",
""
],
[
"Oularabi",
"Abderrahman",
""
],
[
"Anouz",
"Khadija El",
""
],
[
"Demir",
"İlkay",
""
],
[
"Allati",
"Abderrahim El",
""
]
] |
2509.03767
|
Caeley Pittman
|
Caeley V. Pittman, C. C. Espaillat, Zhaohuan Zhu, Thanawuth Thanathibodee, Connor E. Robinson, Nuria Calvet, \'Agnes K\'osp\'al
|
The ODYSSEUS Survey. Using accretion and stellar rotation to reveal the star-disk connection in T Tauri stars
|
Accepted for publication in ApJ. 30 pages, 1 table, 13 figures, 1 figure set. This paper follows Pittman et al. (2025a), located on ADS at https://ui.adsabs.harvard.edu/abs/2025arXiv250701162P/abstract
| null | null | null |
astro-ph.SR astro-ph.EP
|
http://creativecommons.org/licenses/by/4.0/
|
Classical T Tauri stars (CTTS) exhibit strong variability over timescales of minutes to decades. However, much theoretical work assumes that CTTS are in stable spin states. Here, we test expectations for CTTS angular momentum regulation by comparing star and disk rotation. We measure stellar rotation periods and disk corotation radii ($R_{\rm co}$) for 47 CTTS from the HST ULLYSES sample. We compare $R_{\rm co}$ to the magnetospheric truncation radii ($R_{\rm i}$) and show that most CTTS are in the spin-up regime based on model predictions, which may indicate efficient angular momentum loss processes. We find evidence of magnetospheric outflows and episodic accretion, and our observations are consistent with the presence of accretion-powered stellar winds. We confirm predictions that $R_{\rm i}$ is variable over timescales of days, causing some CTTS to cross accretion stability regime boundaries. We characterize light curve morphologies and confirm that our inclined CTTS with $R_{\rm i}\sim R_{\rm co}$ show dipper light curves, consistent with expectations from disk warp models. However, dippers occur at all values of $R_{\rm i}/R_{\rm co}$, suggesting that they do not need to be near the propeller regime. Finally, we show that our measured $R_{\rm i}$ locations are consistent with observed ultra-short-period planet (USP) semi-major axes. If USPs are stable against tidal dissipation, as has been suggested in the literature, then our work provides a plausible USP formation channel. These results show that the star-disk connection produces a large variety of accretion and stellar spin configurations, most of which are likely not in equilibrium.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:23:10 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Pittman",
"Caeley V.",
""
],
[
"Espaillat",
"C. C.",
""
],
[
"Zhu",
"Zhaohuan",
""
],
[
"Thanathibodee",
"Thanawuth",
""
],
[
"Robinson",
"Connor E.",
""
],
[
"Calvet",
"Nuria",
""
],
[
"Kóspál",
"Ágnes",
""
]
] |
2509.03768
|
Koorosh Aslansefat
|
Connor Walker, Koorosh Aslansefat, Mohammad Naveed Akram, Yiannis Papadopoulos
|
RAGuard: A Novel Approach for in-context Safe Retrieval Augmented Generation for LLMs
| null | null | null | null |
cs.AI stat.ML
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Accuracy and safety are paramount in Offshore Wind (OSW) maintenance, yet conventional Large Language Models (LLMs) often fail when confronted with highly specialised or unexpected scenarios. We introduce RAGuard, an enhanced Retrieval-Augmented Generation (RAG) framework that explicitly integrates safety-critical documents alongside technical manuals.By issuing parallel queries to two indices and allocating separate retrieval budgets for knowledge and safety, RAGuard guarantees both technical depth and safety coverage. We further develop a SafetyClamp extension that fetches a larger candidate pool, "hard-clamping" exact slot guarantees to safety. We evaluate across sparse (BM25), dense (Dense Passage Retrieval) and hybrid retrieval paradigms, measuring Technical Recall@K and Safety Recall@K. Both proposed extensions of RAG show an increase in Safety Recall@K from almost 0\% in RAG to more than 50\% in RAGuard, while maintaining Technical Recall above 60\%. These results demonstrate that RAGuard and SafetyClamp have the potential to establish a new standard for integrating safety assurance into LLM-powered decision support in critical maintenance contexts.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:24:17 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Walker",
"Connor",
""
],
[
"Aslansefat",
"Koorosh",
""
],
[
"Akram",
"Mohammad Naveed",
""
],
[
"Papadopoulos",
"Yiannis",
""
]
] |
2509.03769
|
Ying-Cheng Lai
|
Zheng-Meng Zhai, Valerio Lucarini, and Ying-Cheng Lai
|
Deficiency of equation-finding approach to data-driven modeling of dynamical systems
|
6 pages, 3 figures
| null | null | null |
nlin.CD cs.LG math.DS
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Finding the governing equations from data by sparse optimization has become a popular approach to deterministic modeling of dynamical systems. Considering the physical situations where the data can be imperfect due to disturbances and measurement errors, we show that for many chaotic systems, widely used sparse-optimization methods for discovering governing equations produce models that depend sensitively on the measurement procedure, yet all such models generate virtually identical chaotic attractors, leading to a striking limitation that challenges the conventional notion of equation-based modeling in complex dynamical systems. Calculating the Koopman spectra, we find that the different sets of equations agree in their large eigenvalues and the differences begin to appear when the eigenvalues are smaller than an equation-dependent threshold. The results suggest that finding the governing equations of the system and attempting to interpret them physically may lead to misleading conclusions. It would be more useful to work directly with the available data using, e.g., machine-learning methods.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:30:26 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhai",
"Zheng-Meng",
""
],
[
"Lucarini",
"Valerio",
""
],
[
"Lai",
"Ying-Cheng",
""
]
] |
2509.03770
|
Rebecca Jensen-Clem
|
Rebecca Jensen-Clem, Vincent Chambouleyron, Prince Javier, Daren Dillon, Emiel H. Por, Benjamin Calvin, Sylvain Cetre, Rodrigo Amezcua Correa, Tara Crowe, Jordan Diaz, Caleb Dobias, David Doelman, Stephen Eikenberry, J. Fowler, Benjamin L. Gerard, Phil Hinz, Renate Kupke, Ashai Moreno, Tiffany Nguyen, Maissa Salama, Aditya R. Sengupta, Nour Skaf, and Frans Snik
|
The Santa Cruz Extreme AO Lab (SEAL) 2.0: A reflective, multi-wavelength rebuild
|
submitted to SPIE Astronomical Applications 2025, paper number 13627-97
| null | null | null |
astro-ph.IM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The Santa cruz Extreme Adaptive optics Lab (SEAL) is a visible/near-infrared wavelength testbed designed to support technology development for high contrast imaging on large, segmented, ground-based telescopes. SEAL saw first light in 2021 as a transmissive, visible-wavelength AO testbed. In this paper, we present four major upgrades to SEAL: (1) the testbed has been rebuilt with custom off-axis parabolic mirrors, enabling operation in both near-infrared and visible wavelengths; (2) the suite of wavefront sensors now includes a Shack-Hartmann, transmissive four-sided pyramid, vector-Zernike, and, in the muirSEAL testbed, a photonic lantern; (3) the testbed includes a vector-vortex coronagraph and will soon include a hybrid astrophotonic coronagraph; (4) in addition to its original Keck-heritage RTC, SEAL now includes two additional control software packages: Catkit, originally developed for the HiCAT testbed at the Space Telescope Science Institute, and the RTC Compute And Control for Adaptive Optics (CACAO), originally designed for Subaru/SCExAO. We discuss the performance of the testbed after the reflective rebuild and on-going technology development work at SEAL.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:31:15 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Jensen-Clem",
"Rebecca",
""
],
[
"Chambouleyron",
"Vincent",
""
],
[
"Javier",
"Prince",
""
],
[
"Dillon",
"Daren",
""
],
[
"Por",
"Emiel H.",
""
],
[
"Calvin",
"Benjamin",
""
],
[
"Cetre",
"Sylvain",
""
],
[
"Correa",
"Rodrigo Amezcua",
""
],
[
"Crowe",
"Tara",
""
],
[
"Diaz",
"Jordan",
""
],
[
"Dobias",
"Caleb",
""
],
[
"Doelman",
"David",
""
],
[
"Eikenberry",
"Stephen",
""
],
[
"Fowler",
"J.",
""
],
[
"Gerard",
"Benjamin L.",
""
],
[
"Hinz",
"Phil",
""
],
[
"Kupke",
"Renate",
""
],
[
"Moreno",
"Ashai",
""
],
[
"Nguyen",
"Tiffany",
""
],
[
"Salama",
"Maissa",
""
],
[
"Sengupta",
"Aditya R.",
""
],
[
"Skaf",
"Nour",
""
],
[
"Snik",
"Frans",
""
]
] |
2509.03771
|
Brennen Hill
|
Brennen Hill
|
Learning an Adversarial World Model for Automated Curriculum Generation in MARL
| null | null | null | null |
cs.LG cs.AI cs.MA
|
http://creativecommons.org/licenses/by/4.0/
|
World models that infer and predict environmental dynamics are foundational to embodied intelligence. However, their potential is often limited by the finite complexity and implicit biases of hand-crafted training environments. To develop truly generalizable and robust agents, we need environments that scale in complexity alongside the agents learning within them. In this work, we reframe the challenge of environment generation as the problem of learning a goal-conditioned, generative world model. We propose a system where a generative **Attacker** agent learns an implicit world model to synthesize increasingly difficult challenges for a team of cooperative **Defender** agents. The Attacker's objective is not passive prediction, but active, goal-driven interaction: it models and generates world states (i.e., configurations of enemy units) specifically to exploit the Defenders' weaknesses. Concurrently, the embodied Defender team learns a cooperative policy to overcome these generated worlds. This co-evolutionary dynamic creates a self-scaling curriculum where the world model continuously adapts to challenge the decision-making policy of the agents, providing an effectively infinite stream of novel and relevant training scenarios. We demonstrate that this framework leads to the emergence of complex behaviors, such as the world model learning to generate flanking and shielding formations, and the defenders learning coordinated focus-fire and spreading tactics. Our findings position adversarial co-evolution as a powerful method for learning instrumental world models that drive agents toward greater strategic depth and robustness.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:32:39 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Hill",
"Brennen",
""
]
] |
2509.03772
|
Keith Levin
|
Alexander Fuchs-Kreiss, Keith Levin
|
Testing for correlation between network structure and high-dimensional node covariates
| null | null | null | null |
stat.ML cs.LG math.ST stat.TH
|
http://creativecommons.org/licenses/by/4.0/
|
In many application domains, networks are observed with node-level features. In such settings, a common problem is to assess whether or not nodal covariates are correlated with the network structure itself. Here, we present four novel methods for addressing this problem. Two of these are based on a linear model relating node-level covariates to latent node-level variables that drive network structure. The other two are based on applying canonical correlation analysis to the node features and network structure, avoiding the linear modeling assumptions. We provide theoretical guarantees for all four methods when the observed network is generated according to a low-rank latent space model endowed with node-level covariates, which we allow to be high-dimensional. Our methods are computationally cheaper and require fewer modeling assumptions than previous approaches to network dependency testing. We demonstrate and compare the performance of our novel methods on both simulated and real-world data.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:33:17 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Fuchs-Kreiss",
"Alexander",
""
],
[
"Levin",
"Keith",
""
]
] |
2509.03773
|
Kuntal Banerjee
|
Kuntal Banerjee
|
Co-Higgs bundles of Schwarzenberger type and the determinant morphism
|
9 pages
| null | null | null |
math.AG math.DG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We identify images of the determinant morphism of trace-free co-Higgs bundles modeled on rank 2 Schwarzenberger bundles.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:38:17 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Banerjee",
"Kuntal",
""
]
] |
2509.03774
|
Niclas Heinsdorf
|
Niclas Heinsdorf and Marcel Franz
|
Proximitizing altermagnets with conventional superconductors
| null | null | null | null |
cond-mat.supr-con cond-mat.mtrl-sci cond-mat.str-el
|
http://creativecommons.org/licenses/by/4.0/
|
Recent theoretical work highlighted unique properties of superconducting altermagnets, including the wealth of topologically non-trivial phases as well as their potential uses in spintronic applications. Given that no intrinsically superconducting altermagnets have yet been discovered, we study here the possibility of superconducting order induced by proximity effect from a conventional s-wave superconductor. Through symmetry analysis and microscopic modeling we find that interesting superconducting phases can indeed be proximity-induced in a thin altermagnetic film provided that weak Rashba spin-orbit coupling is present at the interface. Surprisingly, the resulting superconductor is generically nodal with a mixed singlet/triplet order parameter and, importantly for applications, capable of generating spin-polarized persistent current. We propose a set of candidate heterostructures with low lattice mismatch suitable to probe these effects experimentally.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:39:45 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Heinsdorf",
"Niclas",
""
],
[
"Franz",
"Marcel",
""
]
] |
2509.03775
|
Sankeerth Durvasula
|
Sankeerth Durvasula, Sharanshangar Muhunthan, Zain Moustafa, Richard Chen, Ruofan Liang, Yushi Guan, Nilesh Ahuja, Nilesh Jain, Selvakumar Panneer, Nandita Vijaykumar
|
ContraGS: Codebook-Condensed and Trainable Gaussian Splatting for Fast, Memory-Efficient Reconstruction
| null | null | null | null |
cs.GR cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
3D Gaussian Splatting (3DGS) is a state-of-art technique to model real-world scenes with high quality and real-time rendering. Typically, a higher quality representation can be achieved by using a large number of 3D Gaussians. However, using large 3D Gaussian counts significantly increases the GPU device memory for storing model parameters. A large model thus requires powerful GPUs with high memory capacities for training and has slower training/rendering latencies due to the inefficiencies of memory access and data movement. In this work, we introduce ContraGS, a method to enable training directly on compressed 3DGS representations without reducing the Gaussian Counts, and thus with a little loss in model quality. ContraGS leverages codebooks to compactly store a set of Gaussian parameter vectors throughout the training process, thereby significantly reducing memory consumption. While codebooks have been demonstrated to be highly effective at compressing fully trained 3DGS models, directly training using codebook representations is an unsolved challenge. ContraGS solves the problem of learning non-differentiable parameters in codebook-compressed representations by posing parameter estimation as a Bayesian inference problem. To this end, ContraGS provides a framework that effectively uses MCMC sampling to sample over a posterior distribution of these compressed representations. With ContraGS, we demonstrate that ContraGS significantly reduces the peak memory during training (on average 3.49X) and accelerated training and rendering (1.36X and 1.88X on average, respectively), while retraining close to state-of-art quality.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:40:17 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Durvasula",
"Sankeerth",
""
],
[
"Muhunthan",
"Sharanshangar",
""
],
[
"Moustafa",
"Zain",
""
],
[
"Chen",
"Richard",
""
],
[
"Liang",
"Ruofan",
""
],
[
"Guan",
"Yushi",
""
],
[
"Ahuja",
"Nilesh",
""
],
[
"Jain",
"Nilesh",
""
],
[
"Panneer",
"Selvakumar",
""
],
[
"Vijaykumar",
"Nandita",
""
]
] |
2509.03776
|
Anbu Venkatachalam
|
Anbu Selvam Venkatachalam, Loren Greenman, Joshua Stallbaumer, Artem Rudenko, Daniel Rolles, and Huynh Van Sa Lam
|
Exploiting correlations in multi-coincidence Coulomb explosion patterns for differentiating molecular structures using machine learning
| null | null | null | null |
physics.chem-ph physics.comp-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Coulomb explosion imaging (CEI) is a powerful technique for capturing the real-time motion of individual atoms during ultrafast photochemical reactions. CEI generates high-dimensional data with naturally embedded correlations that allow mapping the coordinated motion of nuclei in molecules. This enables reliable separation of competing reaction pathways and makes this approach uniquely suited for characterizing weak reaction channels. However, rich information contained in experimental CEI patterns remains largely underexploited due to challenges in visualizing correlations between multiple observables in multi-dimensional parameter space. Here we present a new approach to CEI of intermediate-sized polyatomic molecules, detecting up to eight ionic fragments in coincidence and leveraging machine-learning-based analysis to identify patterns and correlations in the resulting high-dimensional momentum-space data, enabling robust molecular structure identification and differentiation. Our approach provides high-dimensional background-free data encoding exceptionally rich structural information and establishes an automated, scalable framework for extracting insightful information from the data. As a demonstration, we apply this method to image and distinguish dichloroethylene isomers, showcasing its potential for broader applications in molecular imaging. Our results pave the way for channel-specific analysis of ultrafast structural dynamics in chemically relevant systems, particularly for disentangling mixed reaction pathways and detecting contributions from weak channels and minority species.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 23:44:36 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Venkatachalam",
"Anbu Selvam",
""
],
[
"Greenman",
"Loren",
""
],
[
"Stallbaumer",
"Joshua",
""
],
[
"Rudenko",
"Artem",
""
],
[
"Rolles",
"Daniel",
""
],
[
"Lam",
"Huynh Van Sa",
""
]
] |
2509.03777
|
Andrew Graven
|
Andrew J. Graven and Nikolai G. Makarov
|
Quadrature Domains and the Faber Transform
|
60 Pages, 21 Figures
| null | null | null |
math.CV math-ph math.AP math.MP
|
http://creativecommons.org/licenses/by/4.0/
|
We present a framework for reconstructing any simply connected bounded or unbounded, quadrature domain $\Omega$ from its quadrature function $h$. Using the Faber transform, we derive formulae directly relating $h$ to the Riemann map for $\Omega$. Through this approach, we obtain a complete classification of one point quadrature domains with complex charge. We proceed to develop a theory of weighted quadrature domains with respect to weights of the form $\rho_a(w)=|w|^{2(a-1)}$ when $a > 0$ ("power-weighted" quadrature domains) and the limiting case of $a=0$ ("log-weighted" quadrature domains). We obtain Faber transform formulae for reconstructing these from their respective quadrature functions as well. Several examples are presented throughout to illustrate this approach both in the simply connected setting and in the presence of rotational symmetry.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:08:04 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Graven",
"Andrew J.",
""
],
[
"Makarov",
"Nikolai G.",
""
]
] |
2509.03778
|
ZeHong Liao
|
Zehong Liao, Yu Yang, Zepeng Gao, Jun Su, and Long Zhu
|
Mechanism of the quasi-elastic scattering based on the dinuclear system concept
|
8 pages, 5 figures
| null | null | null |
nucl-th
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
A unified description of full reaction channels in low-energy heavy-ion collisions is a great challenge. Although the theoretical models based on the dinuclear system (DNS) concept have been successfully employed in multinucleon transfer (MNT) reactions, the underestimation of the quasi-elastic (QE) channel results in unreliable description of few nucleon transfer, especially for the light reaction systems. In this work, the DNS-sysu model is improved by introducing the impact-parameter-dependent transition probabilities for a unified description of few nucleon and many nucleon transfer in MNT reactions. Extensive experimental data -- including reactions such as 40Ca, 58Ni, 64Ni, 136Xe, and 208Pb + 208Pb -- were compared with the model predictions. The calculated isotopic distributions, mass distributions, and charge distributions show good agreement with experimental measurements. The improved DNS-sysu model enables reasonable characterization and description of the QE/grazing collisions, notably resolving long-standing underestimation in the QE channel.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:20:34 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Liao",
"Zehong",
""
],
[
"Yang",
"Yu",
""
],
[
"Gao",
"Zepeng",
""
],
[
"Su",
"Jun",
""
],
[
"Zhu",
"Long",
""
]
] |
2509.03779
|
Zhiyuan Li
|
Kuang Huang, Zhiyuan Li, Zhidong Zhang, Zhi Zhou
|
H\"older Stable Recovery of the Source in Space-Time Fractional Wave Equations
| null | null | null | null |
math.AP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study the recovery of a spatially dependent source in a one-dimensional space-time fractional wave equation using boundary measurement data collected at a single endpoint. The main challenge arises from the fact that the eigenfunctions of the Dirichlet eigenvalue problem do not form an orthogonal system, due to the presence of a fractional derivative in space. To address this difficulty, we introduce a bi-orthogonal basis for the Mittag-Leffler functions and use it to establish uniqueness and H\"older-type stability results, provided the measurement time is sufficiently large. A Tikhonov regularization method is then employed to numerically solve the inverse source problem. Several numerical examples are presented to demonstrate the accuracy and efficiency of the proposed method and to validate our theoretical findings.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:24:18 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Huang",
"Kuang",
""
],
[
"Li",
"Zhiyuan",
""
],
[
"Zhang",
"Zhidong",
""
],
[
"Zhou",
"Zhi",
""
]
] |
2509.03780
|
John Wentworth
|
John Wentworth, David Lorell
|
Natural Latents: Latent Variables Stable Across Ontologies
| null | null | null | null |
math.PR cs.AI cs.IT cs.LG math.IT
|
http://creativecommons.org/licenses/by/4.0/
|
Suppose two Bayesian agents each learn a generative model of the same environment. We will assume the two have converged on the predictive distribution, i.e. distribution over some observables in the environment, but may have different generative models containing different latent variables. Under what conditions can one agent guarantee that their latents are a function of the other agents latents?
We give simple conditions under which such translation is guaranteed to be possible: the natural latent conditions. We also show that, absent further constraints, these are the most general conditions under which translatability is guaranteed. Crucially for practical application, our theorems are robust to approximation error in the natural latent conditions.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:24:24 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Wentworth",
"John",
""
],
[
"Lorell",
"David",
""
]
] |
2509.03781
|
Brian Vermilyea
|
Brian S. Vermilyea, Michael M. Fogler
|
Plasmons in a network of topological states in twisted bilayer graphene
| null | null | null | null |
cond-mat.mes-hall
|
http://creativecommons.org/licenses/by/4.0/
|
We study surface plasmons in minimally-twisted bilayer graphene that contains a triangular network of partial dislocations (or AB-BA domain walls) hosting one-dimensional electronic states. We calculate plasmon dispersion by solving classical equations of motion for charge dynamics on the network links with impedance boundary conditions at the network nodes. The plasmon band structure is shown to be quasi-periodic in frequency and damped everywhere except at high-symmetry points of the moir\'e Brillouin zone. We compare our network-based formalism with the conventional random phase approximation and discuss when each approach is valid. Calculations of plasmon waves launched by local scatterers are presented to simulate terahertz nano-imaging experiments.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:28:36 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Vermilyea",
"Brian S.",
""
],
[
"Fogler",
"Michael M.",
""
]
] |
2509.03782
|
Fran\c{c}ois Cl\'ement
|
Fran\c{c}ois Cl\'ement
|
Regular Structures in Kronecker Permutations
| null | null | null | null |
math.CO math.NT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Kronecker sequences $(k \alpha \mod 1)_{k=1}^{\infty}$ for some irrational $\alpha > 0$ have played an important role in many areas of mathematics. It is possible to associate to each finite segment $(k \alpha \mod 1)_{k=1}^{n}$ a permutation $\pi \in S_n$ associated with the canonical lifting to two dimensions. We show that these permutations induced by Kronecker sequences based on irrational $\alpha$ are extremely regular for specific choices of $n$ and $\alpha$. In particular, all quadratic irrationals have an infinite number of choices of $n$ that lead to permutations where no cycle has length more than 4.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:28:43 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Clément",
"François",
""
]
] |
2509.03783
|
Nathan Chen
|
Nathan Chen and Olivier Martin
|
A primer on measures of irrationality
|
29 pages; this work grew out of the March 2024 AIM workshop "Degree d points on algebraic surfaces"
| null | null | null |
math.AG
|
http://creativecommons.org/licenses/by/4.0/
|
Measures of irrationality are a numerical way of quantifying how far a given variety is from being rational (or rationally connected, uniruled, etc.). In the last two decades, there has been renewed interest in the study of these invariants. The goal of this expository survey is to summarize known results from the point of view of the Kodaira--Enriques classification of surfaces, highlight recent progress, and discuss a number of open problems and questions.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:31:25 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chen",
"Nathan",
""
],
[
"Martin",
"Olivier",
""
]
] |
2509.03784
|
William Wesley
|
William J. Wesley
|
New bounds for some small multicolor Ramsey numbers
|
Comments welcome
| null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
The Ramsey number $R(G_1,\dots,G_k)$ is the smallest $n$ such that every $k$-coloring of the edges of $K_n$ contains a monochromatic copy of $G_i$ in color $i$. Ramsey numbers are challenging to compute, and few are known exactly. We use Boolean satisfiability (SAT) solvers to search for structured colorings that give lower bounds, and we show $R(K_4,K_4-e,K_4-e) \ge 35$ and $R(K_3,K_4,C_4,C_4) \ge 49$. Moreover, we tighten some recent upper bounds for multicolor Ramsey numbers for cycles and show $R(C_3,C_6,C_6) = R(C_5,C_6,C_6) = 15$. Finally, we enumerate critical graphs for the numbers $R(C_4,K_{1,s})$ and $R(C_6,K_{1,s})$.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:35:53 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Wesley",
"William J.",
""
]
] |
2509.03785
|
Taketo Sano
|
Mikhail Khovanov, Taketo Sano
|
Symmetries of equivariant Khovanov homology
|
30 pages
| null | null |
RIKEN-iTHEMS-Report-25
|
math.QA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study symmetries in equivariant versions of Khovanov homology, which include (i) the construction of an involution $\widehat{\sigma}$ for the $U(2)$-equivariant theory, (ii) an integral lifting $\widehat{\nu}$ of the Shumakovitch operation $\nu$, and (iii) splitting of the $U(1)$- and $U(1)\times U(1)$-equivariant theories generalizing earlier work over $\mathbb{F}_2$. Finally, we relate these structures to the Rasmussen $s$-invariant over an arbitrary field $F$.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:36:16 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Khovanov",
"Mikhail",
""
],
[
"Sano",
"Taketo",
""
]
] |
2509.03786
|
Han Sun
|
Xinxin Wang, Han Sun, Ningzhong Liu, Huiyu Zhou, Yinan Yao
|
SLENet: A Guidance-Enhanced Network for Underwater Camouflaged Object Detection
|
14pages, accepted by PRCV2025
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Underwater Camouflaged Object Detection (UCOD) aims to identify objects that blend seamlessly into underwater environments. This task is critically important to marine ecology. However, it remains largely underexplored and accurate identification is severely hindered by optical distortions, water turbidity, and the complex traits of marine organisms. To address these challenges, we introduce the UCOD task and present DeepCamo, a benchmark dataset designed for this domain. We also propose Semantic Localization and Enhancement Network (SLENet), a novel framework for UCOD. We first benchmark state-of-the-art COD models on DeepCamo to reveal key issues, upon which SLENet is built. In particular, we incorporate Gamma-Asymmetric Enhancement (GAE) module and a Localization Guidance Branch (LGB) to enhance multi-scale feature representation while generating a location map enriched with global semantic information. This map guides the Multi-Scale Supervised Decoder (MSSD) to produce more accurate predictions. Experiments on our DeepCamo dataset and three benchmark COD datasets confirm SLENet's superior performance over SOTA methods, and underscore its high generality for the broader COD task.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:44:32 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Wang",
"Xinxin",
""
],
[
"Sun",
"Han",
""
],
[
"Liu",
"Ningzhong",
""
],
[
"Zhou",
"Huiyu",
""
],
[
"Yao",
"Yinan",
""
]
] |
2509.03787
|
Shakiba Amirshahi
|
Shakiba Amirshahi, Amin Bigdeli, Charles L. A. Clarke, Amira Ghenai
|
Evaluating the Robustness of Retrieval-Augmented Generation to Adversarial Evidence in the Health Domain
| null | null | null | null |
cs.IR cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Retrieval augmented generation (RAG) systems provide a method for factually grounding the responses of a Large Language Model (LLM) by providing retrieved evidence, or context, as support. Guided by this context, RAG systems can reduce hallucinations and expand the ability of LLMs to accurately answer questions outside the scope of their training data. Unfortunately, this design introduces a critical vulnerability: LLMs may absorb and reproduce misinformation present in retrieved evidence. This problem is magnified if retrieved evidence contains adversarial material explicitly intended to promulgate misinformation. This paper presents a systematic evaluation of RAG robustness in the health domain and examines alignment between model outputs and ground-truth answers. We focus on the health domain due to the potential for harm caused by incorrect responses, as well as the availability of evidence-based ground truth for many common health-related questions. We conduct controlled experiments using common health questions, varying both the type and composition of the retrieved documents (helpful, harmful, and adversarial) as well as the framing of the question by the user (consistent, neutral, and inconsistent). Our findings reveal that adversarial documents substantially degrade alignment, but robustness can be preserved when helpful evidence is also present in the retrieval pool. These findings offer actionable insights for designing safer RAG systems in high-stakes domains by highlighting the need for retrieval safeguards. To enable reproducibility and facilitate future research, all experimental results are publicly available in our github repository.
https://github.com/shakibaam/RAG_ROBUSTNESS_EVAL
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:45:58 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Amirshahi",
"Shakiba",
""
],
[
"Bigdeli",
"Amin",
""
],
[
"Clarke",
"Charles L. A.",
""
],
[
"Ghenai",
"Amira",
""
]
] |
2509.03788
|
Kazuhiro Ichihara
|
Kazuhiro Ichihara
|
Finiteness of purely cosmetic fillings
|
7 pages
| null | null | null |
math.GT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
A pair of Dehn fillings on a compact, orientable 3-manifold with a torus boundary is said to be purely cosmetic if the resulting 3-manifolds are orientation-preservingly homeomorphic. In this paper, we show that if the torus boundary is incompressible, then there are only finitely many pairs of purely cosmetic fillings.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:48:02 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Ichihara",
"Kazuhiro",
""
]
] |
2509.03789
|
Muratkhan Abdirash
|
Muratkhan Abdirash and Xiaofan Cui
|
Decentralized Safety-Critical Control of Resilient DC Microgrids with Large-Signal Stability Guarantees
| null | null | null | null |
eess.SY cs.SY
|
http://creativecommons.org/licenses/by/4.0/
|
The increasing penetration of distributed energy resources and power-electronics interfaces in DC microgrids, coupled with rising cyber threats, necessitates primary controllers that are provably safe, cyber-resilient, and practical. The increasing penetration of distributed energy resources and power-electronics interfaces in DC microgrids, coupled with rising cyber threats, necessitates primary controllers that are provably safe, cyber-resilient, and practical. Conventional droop-based methods remain prevalent due to their simplicity, yet their design is largely empirical and conservative, lacking rigorous guarantees. Advanced strategies improve certain aspects, but often sacrifice scalability, robustness, or formal safety. In this work, we propose a Distributed Safety-Critical Controller (DSCC) that systematically integrates global stabilization with formal safety guarantees in a fully decentralized manner. Leveraging control barrier functions and the port-Hamiltonian system theory, the DSCC achieves scalable safe stabilization while preserving real-time implementability. High-fidelity switched-circuit simulations validate the controller's advantages under various contingencies. This framework paves the way for resilient, safety-critical, and scalable control in next-generation DC microgrids.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:51:30 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Abdirash",
"Muratkhan",
""
],
[
"Cui",
"Xiaofan",
""
]
] |
2509.03790
|
Ibne Farabi Shihab
|
Ibne Farabi Shihab, Sanjeda Akter, Anuj Sharma
|
What Fundamental Structure in Reward Functions Enables Efficient Sparse-Reward Learning?
| null | null | null | null |
cs.LG cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
What fundamental properties of reward functions enable efficient sparse-reward reinforcement learning? We address this question through the lens of low-rank structure in reward matrices, showing that such structure induces a sharp transition from exponential to polynomial sample complexity, the first result of this kind for sparse-reward RL. We introduce Policy-Aware Matrix Completion (PAMC), which connects matrix completion theory with reinforcement learning via a new analysis of policy-dependent sampling. Our framework provides: (i) impossibility results for general sparse reward observation, (ii) reward-free representation learning from dynamics, (iii) distribution-free confidence sets via conformal prediction, and (iv) robust completion guarantees that degrade gracefully when low-rank structure is only approximate. Empirically, we conduct a pre-registered evaluation across 100 systematically sampled domains, finding exploitable structure in over half. PAMC improves sample efficiency by factors between 1.6 and 2.1 compared to strong exploration, structured, and representation-learning baselines, while adding only about 20 percent computational overhead.These results establish structural reward learning as a promising new paradigm, with immediate implications for robotics, healthcare, and other safety-critical, sample-expensive applications.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:53:02 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Shihab",
"Ibne Farabi",
""
],
[
"Akter",
"Sanjeda",
""
],
[
"Sharma",
"Anuj",
""
]
] |
2509.03791
|
Saki Imai
|
Saki Imai, Mert \.Inan, Anthony Sicilia, Malihe Alikhani
|
SiLVERScore: Semantically-Aware Embeddings for Sign Language Generation Evaluation
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Evaluating sign language generation is often done through back-translation, where generated signs are first recognized back to text and then compared to a reference using text-based metrics. However, this two-step evaluation pipeline introduces ambiguity: it not only fails to capture the multimodal nature of sign language-such as facial expressions, spatial grammar, and prosody-but also makes it hard to pinpoint whether evaluation errors come from sign generation model or the translation system used to assess it. In this work, we propose SiLVERScore, a novel semantically-aware embedding-based evaluation metric that assesses sign language generation in a joint embedding space. Our contributions include: (1) identifying limitations of existing metrics, (2) introducing SiLVERScore for semantically-aware evaluation, (3) demonstrating its robustness to semantic and prosodic variations, and (4) exploring generalization challenges across datasets. On PHOENIX-14T and CSL-Daily datasets, SiLVERScore achieves near-perfect discrimination between correct and random pairs (ROC AUC = 0.99, overlap < 7%), substantially outperforming traditional metrics.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 00:58:43 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Imai",
"Saki",
""
],
[
"İnan",
"Mert",
""
],
[
"Sicilia",
"Anthony",
""
],
[
"Alikhani",
"Malihe",
""
]
] |
2509.03792
|
Ryo Yonetani
|
Ryo Yonetani and Kotaro Hara
|
Map as a By-product: Collective Landmark Mapping from IMU Data and User-provided Texts in Situated Tasks
|
(c) 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM. This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 9, 3, Article 146 (September 2025), https://doi.org/10.1145/3749455
|
Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 9, 3, Article 146 (September 2025)
|
10.1145/3749455
| null |
cs.HC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents Collective Landmark Mapper, a novel map-as-a-by-product system for generating semantic landmark maps of indoor environments. Consider users engaged in situated tasks that require them to navigate these environments and regularly take notes on their smartphones. Collective Landmark Mapper exploits the smartphone's IMU data and the user's free text input during these tasks to identify a set of landmarks encountered by the user. The identified landmarks are then aggregated across multiple users to generate a unified map representing the positions and semantic information of all landmarks. In developing the proposed system, we focused specifically on retail applications and conducted a formative interview with stakeholders to confirm their practical needs that motivate the map-as-a-byproduct approach. Our user study demonstrates the feasibility of the proposed system and its superior mapping performance in two different setups: creating a product availability map from restocking checklist tasks at a retail store and constructing a room usage map from office inspection tasks, further demonstrating the potential applicability to non-retail applications.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:00:19 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Yonetani",
"Ryo",
""
],
[
"Hara",
"Kotaro",
""
]
] |
2509.03793
|
Prathamesh Devadiga
|
Prathamesh Devadiga, Omkaar Jayadev Shetty, Pooja Agarwal
|
SAMVAD: A Multi-Agent System for Simulating Judicial Deliberation Dynamics in India
| null | null | null | null |
cs.MA cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Understanding the complexities of judicial deliberation is crucial for assessing the efficacy and fairness of a justice system. However, empirical studies of judicial panels are constrained by significant ethical and practical barriers. This paper introduces SAMVAD, an innovative Multi-Agent System (MAS) designed to simulate the deliberation process within the framework of the Indian justice system.
Our system comprises agents representing key judicial roles: a Judge, a Prosecution Counsel, a Defense Counsel, and multiple Adjudicators (simulating a judicial bench), all powered by large language models (LLMs). A primary contribution of this work is the integration of Retrieval-Augmented Generation (RAG), grounded in a domain-specific knowledge base of landmark Indian legal documents, including the Indian Penal Code and the Constitution of India. This RAG functionality enables the Judge and Counsel agents to generate legally sound instructions and arguments, complete with source citations, thereby enhancing both the fidelity and transparency of the simulation.
The Adjudicator agents engage in iterative deliberation rounds, processing case facts, legal instructions, and arguments to reach a consensus-based verdict. We detail the system architecture, agent communication protocols, the RAG pipeline, the simulation workflow, and a comprehensive evaluation plan designed to assess performance, deliberation quality, and outcome consistency.
This work provides a configurable and explainable MAS platform for exploring legal reasoning and group decision-making dynamics in judicial simulations, specifically tailored to the Indian legal context and augmented with verifiable legal grounding via RAG.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:04:44 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Devadiga",
"Prathamesh",
""
],
[
"Shetty",
"Omkaar Jayadev",
""
],
[
"Agarwal",
"Pooja",
""
]
] |
2509.03794
|
Juhun Lee
|
Juhun Lee, Simon S. Woo
|
Fitting Image Diffusion Models on Video Datasets
|
ICCV25 Workshop
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Image diffusion models are trained on independently sampled static images. While this is the bedrock task protocol in generative modeling, capturing the temporal world through the lens of static snapshots is information-deficient by design. This limitation leads to slower convergence, limited distributional coverage, and reduced generalization. In this work, we propose a simple and effective training strategy that leverages the temporal inductive bias present in continuous video frames to improve diffusion training. Notably, the proposed method requires no architectural modification and can be seamlessly integrated into standard diffusion training pipelines. We evaluate our method on the HandCo dataset, where hand-object interactions exhibit dense temporal coherence and subtle variations in finger articulation often result in semantically distinct motions. Empirically, our method accelerates convergence by over 2$\text{x}$ faster and achieves lower FID on both training and validation distributions. It also improves generative diversity by encouraging the model to capture meaningful temporal variations. We further provide an optimization analysis showing that our regularization reduces the gradient variance, which contributes to faster convergence.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:04:54 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Lee",
"Juhun",
""
],
[
"Woo",
"Simon S.",
""
]
] |
2509.03795
|
Hanieh Mousavi
|
Hanieh Mousavi, Anya Jones, Jeff Eldredge
|
Sequential estimation of disturbed aerodynamic flows from sparse measurements via a reduced latent space
| null | null | null | null |
physics.flu-dyn
|
http://creativecommons.org/licenses/by/4.0/
|
This work presents a fast, scalable, and uncertainty-aware methodology for real-time estimation of key aerodynamic states, including instantaneous vorticity fields and aerodynamic loads, during severe gust encounters where strong flow separation and vortex-gust interactions dominate the dynamics. The methodology, trained and tested on high-fidelity simulations of two-dimensional wing-gust encounters with sparse pressure data, combines physics-based ensemble filtering with data-driven surrogate modeling in a learned low-order space. Within this reduced space, latent dynamics and sensor observation operators are learned using neural networks, enabling efficient modeling of complex aerodynamic responses. These components are embedded within a low-rank Ensemble Kalman Filter (LREnKF), yielding a computationally efficient scheme that combines data-driven expressivity with physical interpretability in the original high-dimensional space. Because assimilation occurs entirely within the latent space, updates are fast enough for real-time use, ensuring that aerodynamic states can be continuously estimated from streaming pressure data. The filter corrects predictions predominantly in directions that are both dynamically significant and observable, improving efficiency and reducing spurious adjustments. An observability analysis shows how sensor informativeness evolves during wing-gust interactions, and sensor dropout experiments demonstrate that the LREnKF adaptively re-weights neighboring sensors to preserve estimation quality under degraded sensing.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:08:54 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Mousavi",
"Hanieh",
""
],
[
"Jones",
"Anya",
""
],
[
"Eldredge",
"Jeff",
""
]
] |
2509.03796
|
Joonhwi Joo
|
Joonhwi Joo
|
Selecting the Best Arm in One-Shot Multi-Arm RCTs: The Asymptotic Minimax-Regret Decision Framework for the Best-Population Selection Problem
| null | null | null | null |
econ.EM math.ST stat.ME stat.TH
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We develop a frequentist decision-theoretic framework for selecting the best arm in one-shot, multi-arm randomized controlled trials (RCTs). Our approach characterizes the minimax-regret (MMR) optimal decision rule for any location-family reward distribution with full support. We show that the MMR rule is deterministic, unique, and computationally tractable, as it can be derived by solving the dual problem with nature's least-favorable prior. We then specialize to the case of multivariate normal (MVN) rewards with an arbitrary covariance matrix, and establish the local asymptotic minimaxity of a plug-in version of the rule when only estimated means and covariances are available. This asymptotic MMR (AMMR) procedure maps a covariance-matrix estimate directly into decision boundaries, allowing straightforward implementation in practice. Our analysis highlights a sharp contrast between two-arm and multi-arm designs. With two arms, the empirical success rule ("pick-the-winner") remains MMR-optimal, regardless of the arm-specific variances. By contrast, with three or more arms and heterogeneous variances, the empirical success rule is no longer optimal: the MMR decision boundaries become nonlinear and systematically penalize high-variance arms, requiring stronger evidence to select them. This result underscores that variance plays no role in optimal two-arm comparisons, but it matters critically when more than two options are on the table. Our multi-arm AMMR framework extends classical decision theory to multi-arm RCTs, offering a rigorous foundation and a practical tool for comparing multiple policies simultaneously.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:12:57 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Joo",
"Joonhwi",
""
]
] |
2509.03797
|
Song Guo
|
C. X. Jia, S. Guo, B. Ding, X. H. Zhou, C.X. Yuan, W. Hua J. G. Wang, S. W. Xu, C. M. Petrache, E. A. Lawrie, Y. B. Wu, Y. D. Fang, Y. H. Qiang, Y. Y. Yang, J. B. Ma, J. L. Chen, H. X. Chen, F. Fang, Y. H. Yu, B.F. Lv, F. F. Zeng, Q. B. Zeng, H. Huang, Z. H. Jia, W. Liang, W. Q. Zhang, J. H. Li, J. H. Xu, M. Y. Liu, Y. Zheng, Z. Bai, S. L. Jin, K. Wang, F. F. Duan, G. Yang, G. S. Li, M. L. Liu, Z. Liu, Z. G. Gan, M. Wang, Y. H. Zhang, Y. Q. Liang, Wei Rui, S. Q. Li, H. J. Ong, Y. Li, N. W. Huang, L. J. Liu, and A. Rohilla
|
A high-lying isomer in ^{92}Zr with lifetime modulated by the atomic charge states: a proposed approach for a nuclear gamma-ray laser
| null | null | null | null |
nucl-ex physics.atom-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The nuclides ^{92}Zr are produced and transported by using a radioactive beam line to a lowbackground detection station. After a flight time of about 1.14 {\mu}s, the ions are implanted into a carbon foil, and four {\gamma} rays deexciting the 8+ state in ^{92}Zr are observed in coincidence with the implantation signals within a few nanoseconds. We conjecture that there exists an isomer located slightly above the 8^{+} state in ^{92}Zr. The isomeric lifetime in highly charged states is extended significantly due to the blocking of internal conversion decay channels, enabling its survival over the transportation. During the slowing-down process in the carbon foil, the ^{92}Zr ions capture electron and evolve toward neutral atoms, and consequently the lifetime is restored to a normal short value. Such a high-lying isomer depopulated by a low-energy transition may provide unique opportunity to develop nuclear {\gamma} laser.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:14:04 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Jia",
"C. X.",
""
],
[
"Guo",
"S.",
""
],
[
"Ding",
"B.",
""
],
[
"Zhou",
"X. H.",
""
],
[
"Yuan",
"C. X.",
""
],
[
"Wang",
"W. Hua J. G.",
""
],
[
"Xu",
"S. W.",
""
],
[
"Petrache",
"C. M.",
""
],
[
"Lawrie",
"E. A.",
""
],
[
"Wu",
"Y. B.",
""
],
[
"Fang",
"Y. D.",
""
],
[
"Qiang",
"Y. H.",
""
],
[
"Yang",
"Y. Y.",
""
],
[
"Ma",
"J. B.",
""
],
[
"Chen",
"J. L.",
""
],
[
"Chen",
"H. X.",
""
],
[
"Fang",
"F.",
""
],
[
"Yu",
"Y. H.",
""
],
[
"Lv",
"B. F.",
""
],
[
"Zeng",
"F. F.",
""
],
[
"Zeng",
"Q. B.",
""
],
[
"Huang",
"H.",
""
],
[
"Jia",
"Z. H.",
""
],
[
"Liang",
"W.",
""
],
[
"Zhang",
"W. Q.",
""
],
[
"Li",
"J. H.",
""
],
[
"Xu",
"J. H.",
""
],
[
"Liu",
"M. Y.",
""
],
[
"Zheng",
"Y.",
""
],
[
"Bai",
"Z.",
""
],
[
"Jin",
"S. L.",
""
],
[
"Wang",
"K.",
""
],
[
"Duan",
"F. F.",
""
],
[
"Yang",
"G.",
""
],
[
"Li",
"G. S.",
""
],
[
"Liu",
"M. L.",
""
],
[
"Liu",
"Z.",
""
],
[
"Gan",
"Z. G.",
""
],
[
"Wang",
"M.",
""
],
[
"Zhang",
"Y. H.",
""
],
[
"Liang",
"Y. Q.",
""
],
[
"Rui",
"Wei",
""
],
[
"Li",
"S. Q.",
""
],
[
"Ong",
"H. J.",
""
],
[
"Li",
"Y.",
""
],
[
"Huang",
"N. W.",
""
],
[
"Liu",
"L. J.",
""
],
[
"Rohilla",
"A.",
""
]
] |
2509.03798
|
Dhayaa Anbajagane
|
M. Gatti, D. Anbajagane, C. Chang, D. J. Bacon, J. Prat, M. Adamow, A. Alarcon, M. R. Becker, J. A. Carballo-Bello, N. Chicoine, C. Doux, A. Drlica-Wagner, P. S. Ferguson, D. Gruen, R. A. Gruendl, K. Herron, N. Jeffrey, D. J. James, A. Kov\'acs, C. E. Mart\'inez-V\'azquez, P. Massana, S. Mau, J. McCullough, G. E. Medina, B. Mutlu-Pakdil, N. E. D. No\"el, A. B. Pace, G. Pollina, A. H. Riley, D. J. Sand, L. F. Secco, G. S. Stringfellow, D. Suson, C. Y. Tan, R. Teixeira, E. J. Tollerud, M. A. Troxel, L. Whiteway, A. Zenteno, Z. Zhang
|
DECADE+DES Y3 Weak Lensing Mass Map: A 13,000 deg$^2$ View of Cosmic Structure from 270 Million Galaxies
| null | null | null | null |
astro-ph.CO
|
http://creativecommons.org/licenses/by/4.0/
|
We present the largest galaxy weak lensing mass map of the late-time Universe, reconstructed from 270 million galaxies in the DECADE and DES Year 3 datasets, covering 13,000 square degrees. We validate the map through systematic tests against observational conditions (depth, seeing, etc.), finding the map is statistically consistent with no contamination. The large area covered by the mass map makes it a well-suited tool for cosmological analyses, cross-correlation studies and the identification of large-scale structure features. We demonstrate its potential by detecting cosmic filaments directly from the mass map for the first time and validating them through their association with galaxy clusters selected using the Sunyaev-Zeldovich effect from Planck and ACT DR6.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:14:41 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Gatti",
"M.",
""
],
[
"Anbajagane",
"D.",
""
],
[
"Chang",
"C.",
""
],
[
"Bacon",
"D. J.",
""
],
[
"Prat",
"J.",
""
],
[
"Adamow",
"M.",
""
],
[
"Alarcon",
"A.",
""
],
[
"Becker",
"M. R.",
""
],
[
"Carballo-Bello",
"J. A.",
""
],
[
"Chicoine",
"N.",
""
],
[
"Doux",
"C.",
""
],
[
"Drlica-Wagner",
"A.",
""
],
[
"Ferguson",
"P. S.",
""
],
[
"Gruen",
"D.",
""
],
[
"Gruendl",
"R. A.",
""
],
[
"Herron",
"K.",
""
],
[
"Jeffrey",
"N.",
""
],
[
"James",
"D. J.",
""
],
[
"Kovács",
"A.",
""
],
[
"Martínez-Vázquez",
"C. E.",
""
],
[
"Massana",
"P.",
""
],
[
"Mau",
"S.",
""
],
[
"McCullough",
"J.",
""
],
[
"Medina",
"G. E.",
""
],
[
"Mutlu-Pakdil",
"B.",
""
],
[
"Noël",
"N. E. D.",
""
],
[
"Pace",
"A. B.",
""
],
[
"Pollina",
"G.",
""
],
[
"Riley",
"A. H.",
""
],
[
"Sand",
"D. J.",
""
],
[
"Secco",
"L. F.",
""
],
[
"Stringfellow",
"G. S.",
""
],
[
"Suson",
"D.",
""
],
[
"Tan",
"C. Y.",
""
],
[
"Teixeira",
"R.",
""
],
[
"Tollerud",
"E. J.",
""
],
[
"Troxel",
"M. A.",
""
],
[
"Whiteway",
"L.",
""
],
[
"Zenteno",
"A.",
""
],
[
"Zhang",
"Z.",
""
]
] |
2509.03799
|
Yikan Liu
|
Qingqing Peng, Yikan Liu
|
Energy decay and blow-up of viscoelastic wave equations with polynomial nonlinearity and damping
|
17 pages
| null | null | null |
math.AP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper is concerned with the energy decay and the finite time blow-up of the solution to a viscoelastic wave equation with polynomial nonlinearity and weak damping. We establish explicit and general decay results for the solutions by imposing polynomial conditions on the relaxation function, provided that the initial energy is sufficiently small. Furthermore, we derive an upper bound for the blow-up time when the initial energy is less than the depth of the potential well by utilizing Levine's convexity method. Additionally, we provide a lower bound for the blow-up time if the solution blows up.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:26:43 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Peng",
"Qingqing",
""
],
[
"Liu",
"Yikan",
""
]
] |
2509.03800
|
Yuheng Li
|
Yuheng Li, Yenho Chen, Yuxiang Lai, Jike Zhong, Vanessa Wildman and Xiaofeng Yang
|
MedVista3D: Vision-Language Modeling for Reducing Diagnostic Errors in 3D CT Disease Detection, Understanding and Reporting
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Radiologic diagnostic errors-under-reading errors, inattentional blindness, and communication failures-remain prevalent in clinical practice. These issues often stem from missed localized abnormalities, limited global context, and variability in report language. These challenges are amplified in 3D imaging, where clinicians must examine hundreds of slices per scan. Addressing them requires systems with precise localized detection, global volume-level reasoning, and semantically consistent natural language reporting. However, existing 3D vision-language models are unable to meet all three needs jointly, lacking local-global understanding for spatial reasoning and struggling with the variability and noise of uncurated radiology reports. We present MedVista3D, a multi-scale semantic-enriched vision-language pretraining framework for 3D CT analysis. To enable joint disease detection and holistic interpretation, MedVista3D performs local and global image-text alignment for fine-grained representation learning within full-volume context. To address report variability, we apply language model rewrites and introduce a Radiology Semantic Matching Bank for semantics-aware alignment. MedVista3D achieves state-of-the-art performance on zero-shot disease classification, report retrieval, and medical visual question answering, while transferring well to organ segmentation and prognosis prediction. Code and datasets will be released.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:28:44 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Li",
"Yuheng",
""
],
[
"Chen",
"Yenho",
""
],
[
"Lai",
"Yuxiang",
""
],
[
"Zhong",
"Jike",
""
],
[
"Wildman",
"Vanessa",
""
],
[
"Yang",
"Xiaofeng",
""
]
] |
2509.03801
|
Qiang Cheng
|
Wen-Ting Liu, Shu-Chang Zhao, Qiang Cheng, Qing-Feng Sun
|
Controllable Josephson diode effect, $0$-$\pi$ transition and switch effect in the superconductor/two-dimensional Weyl nodal line semimetal/superconductor junctions
|
10pages,7figures
|
Physical Review B 112 (2025) 054509
|
10.1103/tlqg-268j
| null |
cond-mat.supr-con
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study the Josephson effects in the superconductor/two-dimensional Weyl nodal line semimetal/superconductor junctions using the Green's function method. When the Rashba spin-orbit coupling and an external magnetic field coexist in the semimetal, the symmetries protecting the reciprocity of supercurrent can be broken and the Josephson diode effect with the nonreciprocity of supercurrent can be realized. A high efficiency exceeding $40\%$ can be achieved with the experimentally accessible values of the magnetic field and the Rashba spin-orbit coupling. The diode efficiency can be easily controlled by the direction and magnitude of the field and the strength of the spin-orbit coupling. When the spin-orbit coupling is absent or the external field is absent, the Josephson diode effect vanishes but the current-phase difference relations still show strong dependence on the field or the coupling. The tunable $0$-$\pi$ transition and the switch effect of supercurrent in the junctions can be formed if the direction of the field is rotated or the magnitude of the field and the strength of the coupling are changed. The obtained Josephson diode effect, the $0$-$\pi$ transition and the switch effect of supercurrent are helpful in the design of the quantum devices based on nodal line semimetals.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:35:44 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Liu",
"Wen-Ting",
""
],
[
"Zhao",
"Shu-Chang",
""
],
[
"Cheng",
"Qiang",
""
],
[
"Sun",
"Qing-Feng",
""
]
] |
2509.03802
|
Prateek Sharma
|
Prateek Sharma, Arnav Kumar, Dipayan Datta, Arif Babul, Rishita Das, Konduri Aditya
|
Universal Structure of Turbulent Radiative Mixing Layers
|
6 pages, 4 figures, 1 Table. Comments are welcome. To be submitted. Link to supplementary material is: https://indianinstituteofscience-my.sharepoint.com/:f:/g/personal/prateek_iisc_ac_in/Ess-DCpnwMlGiYc69GlmY5YBLi5C58wVmdPvg9VWziNNhw?e=RWxKNR
| null | null | null |
astro-ph.GA physics.flu-dyn
|
http://creativecommons.org/licenses/by/4.0/
|
Turbulent radiative mixing layers (TRMLs), where shear-driven turbulence between dense and diffuse phases produces intermediate-temperature gas with short cooling times, are ubiquitous in the interstellar and circumgalactic media. Drawing an analogy with Reynolds' decomposition, we perform a quasi-steady-state analysis of TRMLs, separating fields into mean and turbulent components. In the quasi-isobaric TRML, upstream gas cools and compresses before streamwise momentum is fully mixed across the shear layer, which exhibits the expected negative turbulent shear stress $R_{xz}$ that steadily spreads into the cold phase. The thermal pressure reaches a minimum within the TRML, compensated by a positive vertical compressive stress $R_{zz}$. The sum of upstream ram and thermal pressures balances the downstream thermal pressure. Radiative losses are offset by enthalpy dissipation, with a negative turbulent heat flux $Q_{z}$ indicating thermal transport within the TRML. The volume-averaged temperature profile is well fit by a tanh function -- an emergent property that, for the first time, enables a predictive theory for TRMLs consistent with numerical experiments, including the universality of their temperature-dependent emissivity distributions.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:39:01 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Sharma",
"Prateek",
""
],
[
"Kumar",
"Arnav",
""
],
[
"Datta",
"Dipayan",
""
],
[
"Babul",
"Arif",
""
],
[
"Das",
"Rishita",
""
],
[
"Aditya",
"Konduri",
""
]
] |
2509.03803
|
Gao Mengyu
|
Mengyu Gao, Qiulei Dong
|
Causality-guided Prompt Learning for Vision-language Models via Visual Granulation
|
ICCV 2025 Accepted
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Prompt learning has recently attracted much attention for adapting pre-trained vision-language models (e.g., CLIP) to downstream recognition tasks. However, most of the existing CLIP-based prompt learning methods only show a limited ability for handling fine-grained datasets. To address this issue, we propose a causality-guided text prompt learning method via visual granulation for CLIP, called CaPL, where the explored visual granulation technique could construct sets of visual granules for the text prompt to capture subtle discrepancies among different fine-grained classes through casual inference. The CaPL method contains the following two modules: (1) An attribute disentanglement module is proposed to decompose visual features into non-individualized attributes (shared by some classes) and individualized attributes (specific to single classes) using a Brownian Bridge Diffusion Model; (2) A granule learning module is proposed to construct visual granules by integrating the aforementioned attributes for recognition under two causal inference strategies. Thanks to the learned visual granules, more discriminative text prompt is expected to be learned. Extensive experimental results on 15 datasets demonstrate that our CaPL method significantly outperforms the state-of-the-art prompt learning methods, especially on fine-grained datasets.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:40:41 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Gao",
"Mengyu",
""
],
[
"Dong",
"Qiulei",
""
]
] |
2509.03804
|
Md Ragib Shaharear
|
Ad-Deen Mahbub, Md Ragib Shaharear
|
Real-Time Buoyancy Estimation for AUV Simulations Using Convex Hull-Based Submerged Volume Calculation
|
7 pages, 10 figures
|
AQ2UASIM: ADVANCING QUANTITATIVE AND QUALITATIVE SIMULATORS FOR MARINE APPLICATIONS, ICRA(2025)
| null | null |
cs.RO cs.SY eess.SY
|
http://creativecommons.org/licenses/by/4.0/
|
Accurate real-time buoyancy modeling is essential for high-fidelity Autonomous Underwater Vehicle (AUV) simulations, yet NVIDIA Isaac Sim lacks a native buoyancy system, requiring external solutions for precise underwater physics. This paper presents a novel convex hull-based approach to dynamically compute the submerged volume of an AUV in real time. By extracting mesh geometry from the simulation environment and calculating the hull portion intersecting the water level along the z-axis, our method enhances accuracy over traditional geometric approximations. A cross-sectional area extension reduces computational overhead, enabling efficient buoyant force updates that adapt to orientation, depth, and sinusoidal wave fluctuations (+-0.3 m). Tested on a custom AUV design for SAUVC 2025, this approach delivers real-time performance and scalability, improving simulation fidelity for underwater robotics research without precomputed hydrodynamic models.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:42:11 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Mahbub",
"Ad-Deen",
""
],
[
"Shaharear",
"Md Ragib",
""
]
] |
2509.03805
|
Saki Imai
|
Saki Imai, Mert \.Inan, Anthony Sicilia, Malihe Alikhani
|
Measuring How (Not Just Whether) VLMs Build Common Ground
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Large vision language models (VLMs) increasingly claim reasoning skills, yet current benchmarks evaluate them in single-turn or question answering settings. However, grounding is an interactive process in which people gradually develop shared understanding through ongoing communication. We introduce a four-metric suite (grounding efficiency, content alignment, lexical adaptation, and human-likeness) to systematically evaluate VLM performance in interactive grounding contexts. We deploy the suite on 150 self-play sessions of interactive referential games between three proprietary VLMs and compare them with human dyads. All three models diverge from human patterns on at least three metrics, while GPT4o-mini is the closest overall. We find that (i) task success scores do not indicate successful grounding and (ii) high image-utterance alignment does not necessarily predict task success. Our metric suite and findings offer a framework for future research on VLM grounding.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:43:49 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Imai",
"Saki",
""
],
[
"İnan",
"Mert",
""
],
[
"Sicilia",
"Anthony",
""
],
[
"Alikhani",
"Malihe",
""
]
] |
2509.03806
|
Hao Nie
|
Hao Nie, Wei Wang, Peng Xu, Wei Chen, Laurence T. Yang, Mauro Conti, Kaitai Liang
|
Peekaboo, I See Your Queries: Passive Attacks Against DSSE Via Intermittent Observations
| null | null | null | null |
cs.CR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Dynamic Searchable Symmetric Encryption (DSSE) allows secure searches over a dynamic encrypted database but suffers from inherent information leakage. Existing passive attacks against DSSE rely on persistent leakage monitoring to infer leakage patterns, whereas this work targets intermittent observation - a more practical threat model. We propose Peekaboo - a new universal attack framework - and the core design relies on inferring the search pattern and further combining it with auxiliary knowledge and other leakage. We instantiate Peekaboo over the SOTA attacks, Sap (USENIX' 21) and Jigsaw (USENIX' 24), to derive their "+" variants (Sap+ and Jigsaw+). Extensive experiments demonstrate that our design achieves >0.9 adjusted rand index for search pattern recovery and 90% query accuracy vs. FMA's 30% (CCS' 23). Peekaboo's accuracy scales with observation rounds and the number of observed queries but also it resists SOTA countermeasures, with >40% accuracy against file size padding and >80% against obfuscation.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:47:22 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Nie",
"Hao",
""
],
[
"Wang",
"Wei",
""
],
[
"Xu",
"Peng",
""
],
[
"Chen",
"Wei",
""
],
[
"Yang",
"Laurence T.",
""
],
[
"Conti",
"Mauro",
""
],
[
"Liang",
"Kaitai",
""
]
] |
2509.03807
|
Wei Wang
|
Junhui Li, Chengbin Feng, Zhiwei Yang, Qi Mo and Wei Wang
|
BIDO: A Unified Approach to Address Obfuscation and Concept Drift Challenges in Image-based Malware Detection
| null | null | null | null |
cs.CR cs.SE
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
To identify malicious Android applications, various malware detection techniques have been proposed. Among them, image-based approaches are considered potential alternatives due to their efficiency and scalability. Recent studies have reported that these approaches suffer significant performance declines when confronted with obfuscation or concept drift. However, existing solutions often treat these two challenges as different problems, offering independent solutions. These techniques overlook the fact that both challenges share a common statistical root, out-of-distribution, and research from this perspective remains limited. In response, we propose BIDO, a hybrid image-based malware detector designed to enhance robustness against both obfuscation and concept drift simultaneously. Specifically, to improve the discriminative power of image features, we introduce a local feature selection module that identifies informative subregions within malware images. Second, to enhance feature robustness, we model pairwise cross-modal dependencies in an outer product space, enabling the extraction of stable co-occurrence patterns. Third, to ensure feature compactness, we design a learnable metric that pulls samples with identical labels closer while pushing apart those with different labels, regardless of obfuscation or concept drift. Extensive experiments on the real-world datasets demonstrate that BIDO significantly outperforms existing baselines, achieving higher robustness against both concept drift and obfuscation. The source code is available at: https://github.com/whatishope/BIDO/.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:48:03 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Li",
"Junhui",
""
],
[
"Feng",
"Chengbin",
""
],
[
"Yang",
"Zhiwei",
""
],
[
"Mo",
"Qi",
""
],
[
"Wang",
"Wei",
""
]
] |
2509.03808
|
Weidong Hao
|
Huanan Li, Rui Fan, Juntao Guan, Weidong Hao, Lai Rui, Tong Wu, Yikai Wang, Lin Gu
|
EGTM: Event-guided Efficient Turbulence Mitigation
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Turbulence mitigation (TM) aims to remove the stochastic distortions and blurs introduced by atmospheric turbulence into frame cameras. Existing state-of-the-art deep-learning TM methods extract turbulence cues from multiple degraded frames to find the so-called "lucky'', not distorted patch, for "lucky fusion''. However, it requires high-capacity network to learn from coarse-grained turbulence dynamics between synchronous frames with limited frame-rate, thus fall short in computational and storage efficiency. Event cameras, with microsecond-level temporal resolution, have the potential to fundamentally address this bottleneck with efficient sparse and asynchronous imaging mechanism. In light of this, we (i) present the fundamental \textbf{``event-lucky insight''} to reveal the correlation between turbulence distortions and inverse spatiotemporal distribution of event streams. Then, build upon this insight, we (ii) propose a novel EGTM framework that extracts pixel-level reliable turbulence-free guidance from the explicit but noisy turbulent events for temporal lucky fusion. Moreover, we (iii) build the first turbulence data acquisition system to contribute the first real-world event-driven TM dataset. Extensive experimental results demonstrate that our approach significantly surpass the existing SOTA TM method by 710 times, 214 times and 224 times in model size, inference latency and model complexity respectively, while achieving the state-of-the-art in restoration quality (+0.94 PSNR and +0.08 SSIM) on our real-world EGTM dataset. This demonstrating the great efficiency merit of introducing event modality into TM task. Demo code and data have been uploaded in supplementary material and will be released once accepted.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:49:13 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Li",
"Huanan",
""
],
[
"Fan",
"Rui",
""
],
[
"Guan",
"Juntao",
""
],
[
"Hao",
"Weidong",
""
],
[
"Rui",
"Lai",
""
],
[
"Wu",
"Tong",
""
],
[
"Wang",
"Yikai",
""
],
[
"Gu",
"Lin",
""
]
] |
2509.03809
|
Jiaxin Guo
|
Jiaxin Guo, Daimeng Wei, Yuanchang Luo, Xiaoyu Chen, Zhanglin Wu, Huan Yang, Hengchao Shang, Zongyao Li, Zhiqiang Rao, Jinlong Yang and Hao Yang
|
Align-then-Slide: A complete evaluation framework for Ultra-Long Document-Level Machine Translation
|
under preview
| null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Large language models (LLMs) have ushered in a new era for document-level machine translation (\textit{doc}-mt), yet their whole-document outputs challenge existing evaluation methods that assume sentence-by-sentence alignment. We introduce \textit{\textbf{Align-then-Slide}}, a complete evaluation framework for ultra-long doc-mt. In the Align stage, we automatically infer sentence-level source-target correspondences and rebuild the target to match the source sentence number, resolving omissions and many-to-one/one-to-many mappings. In the n-Chunk Sliding Evaluate stage, we calculate averaged metric scores under 1-, 2-, 3- and 4-chunk for multi-granularity assessment. Experiments on the WMT benchmark show a Pearson correlation of 0.929 between our method with expert MQM rankings. On a newly curated real-world test set, our method again aligns closely with human judgments. Furthermore, preference data produced by Align-then-Slide enables effective CPO training and its direct use as a reward model for GRPO, both yielding translations preferred over a vanilla SFT baseline. The results validate our framework as an accurate, robust, and actionable evaluation tool for doc-mt systems.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:50:20 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Guo",
"Jiaxin",
""
],
[
"Wei",
"Daimeng",
""
],
[
"Luo",
"Yuanchang",
""
],
[
"Chen",
"Xiaoyu",
""
],
[
"Wu",
"Zhanglin",
""
],
[
"Yang",
"Huan",
""
],
[
"Shang",
"Hengchao",
""
],
[
"Li",
"Zongyao",
""
],
[
"Rao",
"Zhiqiang",
""
],
[
"Yang",
"Jinlong",
""
],
[
"Yang",
"Hao",
""
]
] |
2509.03810
|
Xiannan Huang
|
Xiannan Huang, Shuhan Qiu, Jiayuan Du, Chao Yang
|
Online time series prediction using feature adjustment
| null | null | null | null |
cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Time series forecasting is of significant importance across various domains. However, it faces significant challenges due to distribution shift. This issue becomes particularly pronounced in online deployment scenarios where data arrives sequentially, requiring models to adapt continually to evolving patterns. Current time series online learning methods focus on two main aspects: selecting suitable parameters to update (e.g., final layer weights or adapter modules) and devising suitable update strategies (e.g., using recent batches, replay buffers, or averaged gradients). We challenge the conventional parameter selection approach, proposing that distribution shifts stem from changes in underlying latent factors influencing the data. Consequently, updating the feature representations of these latent factors may be more effective. To address the critical problem of delayed feedback in multi-step forecasting (where true values arrive much later than predictions), we introduce ADAPT-Z (Automatic Delta Adjustment via Persistent Tracking in Z-space). ADAPT-Z utilizes an adapter module that leverages current feature representations combined with historical gradient information to enable robust parameter updates despite the delay. Extensive experiments demonstrate that our method consistently outperforms standard base models without adaptation and surpasses state-of-the-art online learning approaches across multiple datasets. The code is available at https://github.com/xiannanhuang/ADAPT-Z.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:54:48 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Huang",
"Xiannan",
""
],
[
"Qiu",
"Shuhan",
""
],
[
"Du",
"Jiayuan",
""
],
[
"Yang",
"Chao",
""
]
] |
2509.03811
|
Jiaheng Yin
|
Yongzhi Qi, Jiaheng Yin, Jianshen Zhang, Dongyang Geng, Zhengyu Chen, Hao Hu, Wei Qi, Zuo-Jun Max Shen
|
Leveraging LLM-Based Agents for Intelligent Supply Chain Planning
| null | null | null | null |
cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In supply chain management, planning is a critical concept. The movement of physical products across different categories, from suppliers to warehouse management, to sales, and logistics transporting them to customers, entails the involvement of many entities. It covers various aspects such as demand forecasting, inventory management, sales operations, and replenishment. How to collect relevant data from an e-commerce platform's perspective, formulate long-term plans, and dynamically adjust them based on environmental changes, while ensuring interpretability, efficiency, and reliability, is a practical and challenging problem. In recent years, the development of AI technologies, especially the rapid progress of large language models, has provided new tools to address real-world issues. In this work, we construct a Supply Chain Planning Agent (SCPA) framework that can understand domain knowledge, comprehend the operator's needs, decompose tasks, leverage or create new tools, and return evidence-based planning reports. We deploy this framework in JD.com's real-world scenario, demonstrating the feasibility of LLM-agent applications in the supply chain. It effectively reduced labor and improved accuracy, stock availability, and other key metrics.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:55:58 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Qi",
"Yongzhi",
""
],
[
"Yin",
"Jiaheng",
""
],
[
"Zhang",
"Jianshen",
""
],
[
"Geng",
"Dongyang",
""
],
[
"Chen",
"Zhengyu",
""
],
[
"Hu",
"Hao",
""
],
[
"Qi",
"Wei",
""
],
[
"Shen",
"Zuo-Jun Max",
""
]
] |
2509.03812
|
Shadeeb Hossain
|
Shadeeb Hossain, Natalie Sommer, Neda Adib
|
Exploring the Integration of Extended Reality and Artificial Intelligence (AI) for Remote STEM Education and Assessment
|
9 pages, 5 figures, 1 table
| null | null | null |
cs.HC
|
http://creativecommons.org/licenses/by/4.0/
|
This paper presents a dynamic gamification architecture for an Extended Reality Artificial Intelligence virtual training environment designed to enhance STEM education through immersive adaptive, and kinesthetic learning. The proposed system can be introduced in four phases: Introduction Phase, Component Development Phase, Fault Introduction and Correction Phase and Generative AI XR scenarios Phase. Security and privacy are discussed via a defense-in-depth approach spanning client, middleware, and backend layers, incorporating AES 256 encryption, multi-factor authentication, role-based access control and GDPR or FERPA compliance. Risks such as sensor exploitation, perceptual manipulation, and virtual physical harm are identified, with mitigation strategies embedded at the design stage. Potential barriers to large scale adoption-including technical complexity, cost of deployment, and need for cybersecurity expertise are discussed.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:57:45 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Hossain",
"Shadeeb",
""
],
[
"Sommer",
"Natalie",
""
],
[
"Adib",
"Neda",
""
]
] |
2509.03813
|
Mohammed Eltayeb
|
Parth Ashokbhai Shiroya, Swarnagowri Shashidhar, Amod Ashtekar, Krishna Aindrila Kar, Rafaela Lomboy, Dalton Davis, and Mohammed E. Eltayeb
|
Machine Learning for LiDAR-Based Indoor Surface Classification in Intelligent Wireless Environments
| null | null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Reliable connectivity in millimeter-wave (mmWave) and sub-terahertz (sub-THz) networks depends on reflections from surrounding surfaces, as high-frequency signals are highly vulnerable to blockage. The scattering behavior of a surface is determined not only by material permittivity but also by roughness, which governs whether energy remains in the specular direction or is diffusely scattered. This paper presents a LiDAR-driven machine learning framework for classifying indoor surfaces into semi-specular and low-specular categories, using optical reflectivity as a proxy for electromagnetic scattering behavior. A dataset of over 78,000 points from 15 representative indoor materials was collected and partitioned into 3 cm x 3 cm patches to enable classification from partial views. Patch-level features capturing geometry and intensity, including elevation angle, natural-log-scaled intensity, and max-to-mean ratio, were extracted and used to train Random Forest, XGBoost, and neural network classifiers. Results show that ensemble tree-based models consistently provide the best trade-off between accuracy and robustness, confirming that LiDAR-derived features capture roughness-induced scattering effects. The proposed framework enables the generation of scatter aware environment maps and digital twins, supporting adaptive beam management, blockage recovery, and environment-aware connectivity in next-generation networks.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 01:58:18 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Shiroya",
"Parth Ashokbhai",
""
],
[
"Shashidhar",
"Swarnagowri",
""
],
[
"Ashtekar",
"Amod",
""
],
[
"Kar",
"Krishna Aindrila",
""
],
[
"Lomboy",
"Rafaela",
""
],
[
"Davis",
"Dalton",
""
],
[
"Eltayeb",
"Mohammed E.",
""
]
] |
2509.03814
|
Jude Pereira
|
Jude Pereira, Tanmay Vachaspati
|
Electric flux tube solutions in SU(3) gauge theory
|
9 pages, 2 figures
| null | null | null |
hep-th hep-ph
|
http://creativecommons.org/licenses/by/4.0/
|
We consider electric flux tube solutions in SU(3) gauge theory with scalar fields in the fundamental representation. Such solutions can possibly be constructed in two classes, corresponding to the two maximally commuting generators $\lambda_3$ and $\lambda_8$ of SU(3). We successfully construct the 3-type solution with two scalar fields and show that it is immune to decay through Schwinger pair production of gauge bosons but we have not been able to construct an 8-type solution.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:01:00 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Pereira",
"Jude",
""
],
[
"Vachaspati",
"Tanmay",
""
]
] |
2509.03815
|
Kai Zhang
|
Kai Zhang, Situ Wang, Linghang Kong, Fang Zhang, Zhengfeng Ji and Jianxin Chen
|
Learning Neural Decoding with Parallelism and Self-Coordination for Quantum Error Correction
|
21 pages, 11 figures
| null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Fast, reliable decoders are pivotal components for enabling fault-tolerant quantum computation. Neural network decoders like AlphaQubit have demonstrated significant potential, achieving higher accuracy than traditional human-designed decoding algorithms. However, existing implementations of neural network decoders lack the parallelism required to decode the syndrome stream generated by a superconducting logical qubit in real time. Moreover, integrating AlphaQubit with sliding window-based parallel decoding schemes presents non-trivial challenges: AlphaQubit is trained solely to output a single bit corresponding to the global logical correction for an entire memory experiment, rather than local physical corrections that can be easily integrated.
We address this issue by training a recurrent, transformer-based neural network specifically tailored for sliding-window decoding. While our network still outputs a single bit per window, we derive training labels from a consistent set of local corrections and train on various types of decoding windows simultaneously. This approach enables the network to self-coordinate across neighboring windows, facilitating high-accuracy parallel decoding of arbitrarily long memory experiments. As a result, we resolve the throughput limitation that previously prohibited the application of AlphaQubit-type decoders in fault-tolerant quantum computation.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:02:27 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhang",
"Kai",
""
],
[
"Wang",
"Situ",
""
],
[
"Kong",
"Linghang",
""
],
[
"Zhang",
"Fang",
""
],
[
"Ji",
"Zhengfeng",
""
],
[
"Chen",
"Jianxin",
""
]
] |
2509.03816
|
Aman Gupta
|
Aman Gupta, Aditi Sheshadri, Sujit Roy, Johannes Schmude, Vishal Gaur, Wei Ji Leong, Manil Maskey, Rahul Ramachandran
|
Finetuning AI Foundation Models to Develop Subgrid-Scale Parameterizations: A Case Study on Atmospheric Gravity Waves
| null | null | null | null |
physics.ao-ph cs.LG
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Global climate models parameterize a range of atmospheric-oceanic processes like gravity waves, clouds, moist convection, and turbulence that cannot be sufficiently resolved. These subgrid-scale closures for unresolved processes are a leading source of model uncertainty. Here, we present a new approach to developing machine learning parameterizations of small-scale climate processes by fine-tuning a pre-trained AI foundation model (FM). FMs are largely unexplored in climate research. A pre-trained encoder-decoder from a 2.3 billion parameter FM (NASA and IBM Research's Prithvi WxC) -- which contains a latent probabilistic representation of atmospheric evolution -- is fine-tuned (or reused) to create a deep learning parameterization for atmospheric gravity waves (GWs). The parameterization captures GW effects for a coarse-resolution climate model by learning the fluxes from an atmospheric reanalysis with 10 times finer resolution. A comparison of monthly averages and instantaneous evolution with a machine learning model baseline (an Attention U-Net) reveals superior predictive performance of the FM parameterization throughout the atmosphere, even in regions excluded from pre-training. This performance boost is quantified using the Hellinger distance, which is 0.11 for the baseline and 0.06 for the fine-tuned model. Our findings emphasize the versatility and reusability of FMs, which could be used to accomplish a range of atmosphere- and climate-related applications, leading the way for the creation of observations-driven and physically accurate parameterizations for more earth-system processes.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:05:54 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Gupta",
"Aman",
""
],
[
"Sheshadri",
"Aditi",
""
],
[
"Roy",
"Sujit",
""
],
[
"Schmude",
"Johannes",
""
],
[
"Gaur",
"Vishal",
""
],
[
"Leong",
"Wei Ji",
""
],
[
"Maskey",
"Manil",
""
],
[
"Ramachandran",
"Rahul",
""
]
] |
2509.03817
|
Wei Yang
|
Wei Yang, Jesse Thomason
|
Learning to Deliberate: Meta-policy Collaboration for Agentic LLMs with Multi-agent Reinforcement Learning
| null | null | null | null |
cs.AI cs.MA
|
http://creativecommons.org/licenses/by/4.0/
|
Multi-agent systems of large language models (LLMs) show promise for complex reasoning, but their effectiveness is often limited by fixed collaboration protocols. These frameworks typically focus on macro-level orchestration while overlooking agents' internal deliberative capabilities. This critical meta-cognitive blindspot treats agents as passive executors unable to adapt their strategy based on internal cognitive states like uncertainty or confidence. We introduce the Meta-Policy Deliberation Framework (MPDF), where agents learn a decentralized policy over a set of high-level meta-cognitive actions: Persist, Refine, and Concede. To overcome the instability of traditional policy gradients in this setting, we develop SoftRankPO, a novel reinforcement learning algorithm. SoftRankPO stabilizes training by shaping advantages based on the rank of rewards mapped through smooth normal quantiles, making the learning process robust to reward variance. Experiments show that MPDF with SoftRankPO achieves a a 4-5% absolute gain in average accuracy across five mathematical and general reasoning benchmarks compared to six state-of-the-art heuristic and learning-based multi-agent reasoning algorithms. Our work presents a paradigm for learning adaptive, meta-cognitive policies for multi-agent LLM systems, shifting the focus from designing fixed protocols to learning dynamic, deliberative strategies.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:06:06 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Yang",
"Wei",
""
],
[
"Thomason",
"Jesse",
""
]
] |
2509.03818
|
Sherwan Abdullah
|
Sherwan Jalal Abdullah, Sravan Reddy Chintareddy, Victor S. Frost, Shawn Keshmiri, Morteza Hashemi
|
A Versatile and Programmable UAV Platform for Radio Access Network and End-to-End Cellular Measurements
| null | null | null | null |
cs.NI cs.SY eess.SY
|
http://creativecommons.org/licenses/by/4.0/
|
In this work, we develop a measurement platform to capture mobile network performance metrics including coverage and quality of service in regions where conventional coverage testing approaches are frequently time-intensive, labor-demanding, and occasionally hazardous. Traditionally, crowd-sourcing methods are used to collect cellular network performance metrics. However, these approaches are inadequate in rural areas due to low-density population, and difficult terrain. The platform described here is a UAV-based and is designed to investigate the mobile network performance through aerial operations and gather Radio Access Network (RAN) signal alongside end-to-end network performance metrics. Our platform gathers metrics through the integration of an onboard computation unit and commercial off-the-shelf cellular modem. The gathered data are subsequently analyzed and displayed using geospatial mapping utilities and statistical techniques to deliver key observations on cellular network performance. Experimental results showed that the received signal power improves at higher altitudes due to enhanced line-of-sight (LoS) conditions as expected. However, the signal quality degrades as a result of increased interference from neighboring cells. The analysis reveals that for most of the geographic area covered in the initial experiments the system maintained acceptable signal quality, with adequate throughput performance for both uplink and downlink communications, while maintaining satisfactory round-trip time characteristics. Notably, the experiment showed that a strong radio signal metric for a given cell does not necessarily translate to consistent spatial coverage across the tested region.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:06:06 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Abdullah",
"Sherwan Jalal",
""
],
[
"Chintareddy",
"Sravan Reddy",
""
],
[
"Frost",
"Victor S.",
""
],
[
"Keshmiri",
"Shawn",
""
],
[
"Hashemi",
"Morteza",
""
]
] |
2509.03819
|
Pablo Rivas
|
Meghan Bibb, Pablo Rivas, and Mahee Tayba
|
Predicting Traffic Accident Severity with Deep Neural Networks
|
The 17th International Conference on Data Science (ICDATA 2021)
| null | null | null |
cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Traffic accidents can be studied to mitigate the risk of further events. Recent advances in machine learning have provided an alternative way to study data associated with traffic accidents. New models achieve good generalization and high predictive power over imbalanced data. In this research, we study neural network-based models on data related to traffic accidents. We begin analyzing relative feature colinearity and unsupervised dimensionality reduction through autoencoders, followed by a dense network. The features are related to traffic accident data and the target is to classify accident severity. Our experiments show cross-validated results of up to 92% accuracy when classifying accident severity using the proposed deep neural network.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:08:44 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Bibb",
"Meghan",
""
],
[
"Rivas",
"Pablo",
""
],
[
"Tayba",
"Mahee",
""
]
] |
2509.03820
|
Gus Schrader
|
Gus Schrader, Alexander Shapiro
|
The algebraic modular functor conjecture in type $A_n$ quantum Teichm\"uller theory
|
145 pages, 65 figures. Comments welcomed!
| null | null | null |
math.QA math.RT
|
http://creativecommons.org/licenses/by/4.0/
|
Fock and Goncharov introduced a quantization of higher Teichm\"uller theory using cluster Poisson varieties and their noncommutative deformations, associating to a complex semisimple Lie group $G$ and a marked surface $S$ a quantum algebra $\mathbb{L}_{G,S}$ equipped with an action of the surface mapping class group. They conjectured that these quantizations form an algebraic analog of a modular functor: cutting a surface along a simple closed curve should correspond to a canonical gluing isomorphism for the associated algebras. In this paper we prove this conjecture for $G = \mathrm{PGL}_{n+1}$. Our approach requires two extensions of the Fock-Goncharov framework: (1) enhanced moduli spaces incorporating additional boundary data, providing algebro-geometric analogs of Fenchel-Nielsen twist coordinates; and (2) the residue universal Laurent ring, a refinement of the quantum universal Laurent ring obtained by localizing and imposing residue conditions. Using these tools, we construct canonical cutting isomorphisms that are equivariant under mapping class group actions and suffice to reconstruct the entire algebra $\mathbb{L}_{G,S}$ from data associated to the cut surface.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:11:30 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Schrader",
"Gus",
""
],
[
"Shapiro",
"Alexander",
""
]
] |
2509.03821
|
Rui Zhao
|
Rui Zhao, Muhammad Shoaib, Viet Tung Hoang, and Wajih Ul Hassan
|
Rethinking Tamper-Evident Logging: A High-Performance, Co-Designed Auditing System
|
This paper has been accepted to the ACM Conference on Computer and Communications Security (CCS) 2025
|
Proceedings of the 2025 ACM SIGSAC Conference on Computer and Communications Security
|
10.1145/3719027.3765024
| null |
cs.CR
|
http://creativecommons.org/licenses/by/4.0/
|
Existing tamper-evident logging systems suffer from high overhead and severe data loss in high-load settings, yet only provide coarse-grained tamper detection. Moreover, installing such systems requires recompiling kernel code. To address these challenges, we present Nitro, a high-performance, tamper-evident audit logging system that supports fine-grained detection of log tampering. Even better, our system avoids kernel recompilation by using the eBPF technology. To formally justify the security of Nitro, we provide a new definitional framework for logging systems, and give a practical cryptographic construction meeting this new goal. Unlike prior work that focus only on the cryptographic processing, we codesign the cryptographic part with the pre- and post-processing of the logs to exploit all system-level optimizations. Our evaluations demonstrate Nitro's superior performance, achieving 10X-25X improvements in high-stress conditions and 2X-10X in real-world scenarios while maintaining near-zero data loss. We also provide an advanced variant, Nitro-R that introduces in-kernel log reduction techniques to reduce runtime overhead even further.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:12:40 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhao",
"Rui",
""
],
[
"Shoaib",
"Muhammad",
""
],
[
"Hoang",
"Viet Tung",
""
],
[
"Hassan",
"Wajih Ul",
""
]
] |
2509.03822
|
Xingjian Yi
|
Xing-Jian Yi, Yue Mao, Xiancong Lu, Qing-Feng Sun
|
Spin Splitting Nernst Effect in Altermagnet
| null | null | null | null |
cond-mat.mes-hall
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Altermagnet is a distinctive magnet phase, which has spin-split energy band but with zero net magnetic moment. In this paper, we propose that altermagnet behaves spin splitting Nernst effect: Under a longitudinal temperature gradient, the electrons with opposite spins tend to split oppositely in the transverse direction, thus generating a transverse spin current. The spin splitting Nernst effect is understood from the contribution of the longitudinal wave vector to the transverse group velocity. Using the nonequilibrium Green's function method, we calculate the spin-dependent transmission coefficient in the four-terminal altermagnet device. From the spin-dependent transmission coefficient, the nonzero transverse spin current from longitudinal temperature gradient is obtained, and the spin splitting Nernst effect is verified. We systematically study the parameter dependence of the spin splitting Nernst effect, while also performing symmetry analysis. The spin splitting Nernst effect can be easily regulated by Fermi surface energy, temperature, transport direction, and system size. Furthermore, in altermagnet, the $xy$-response and $yx$-response spin splitting Nernst coefficients are equal with $N_{s,xy}=N_{s,yx}$, different from the conventional spin Nernst effect where they are opposite. Meanwhile, the spin splitting Nernst effect require neither spin-orbit coupling nor net magnetism.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:15:21 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Yi",
"Xing-Jian",
""
],
[
"Mao",
"Yue",
""
],
[
"Lu",
"Xiancong",
""
],
[
"Sun",
"Qing-Feng",
""
]
] |
2509.03823
|
Jingsong Chai
|
Jingsong Chai
|
On parabolic induction and Jacquet modules over p-adic group
| null | null | null | null |
math.RT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this note, using tensor products with appropriate bimodules over Hecke algebras, we uniformly describe parabolic induction and Jacquet module. We also recover a result of Loke and Przebinda on construction of big theta lift in local theta correspondenc using similar description.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:16:04 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chai",
"Jingsong",
""
]
] |
2509.03824
|
Hidekazu Yoshioka
|
H. Yoshioka
|
A minimization principle behind the diffusion bridge of diurnal fish migration
| null | null | null | null |
math.OC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Fish migration is a mass movement that affects the hydrosphere and ecosystems. While it occurs on multiple temporal scales, including daily and intraday fluctuations, the latter remains less studied. In this study, for a stochastic differential equation model of the intraday unit-time fish count at a fixed observation point, we demonstrate that the model can be derived from a minimization problem in the form of a stochastic control problem. The control problem assumes the form of the Schr\"odinger Bridge but differs from classical formulations by involving a degenerate diffusion process and an objective function with a novel time-dependent weight coefficient. The well-posedness of the control problem and its solution are discussed in detail, using a penalized formulation. The proposed theory is applied to juvenile upstream migration events of the diadromous fish species Plecoglossus altivelis altivelis commonly called Ayu in Japan. We also conduct sensitivity analysis of the models identified from real data.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:17:49 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Yoshioka",
"H.",
""
]
] |
2509.03825
|
Jeunghoon Lee
|
Jeunghoon Lee
|
Sensor placement for sparse force reconstruction
| null |
Mechanical Systems and Signal Processing 2025
| null | null |
eess.SP
|
http://creativecommons.org/licenses/by/4.0/
|
The present study proposes a Gram-matrix-based sensor placement strategy for sparse force reconstruction in the frequency domain. A modal decomposition of the Gram matrix reveals that its structure is dominated by a few modes near the target frequency, and that each modal contribution reflects the spatial correlation of the corresponding mode shape. This suggests that placing sensors near nodal regions where spatial correlation is low can reduce coherence in the frequency response function (FRF) matrix and improve force reconstruction accuracy. To translate the physical insight into a practical design framework, a greedy algorithm is proposed to select sensor locations that minimize the off-diagonal energy of the Gram matrix. Numerical simulations and experimental validations demonstrate that the proposed method yields robust and accurate force estimation, outperforming heuristic sensor layouts.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:19:35 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Lee",
"Jeunghoon",
""
]
] |
2509.03826
|
Tony Wells Dr
|
Tony Wells, B. C. Addison, R.A. Wittenmyer, Duncan J. Wright, Tyler R. Fairnington, Jason Dittmann, Jonathan Horner, Stephen R. Kane, John Kielkopf, Peter Plavchan and Avi Shporer
|
The Spin-orbit alignment of two short period eclipsing binary systems
| null | null | null | null |
astro-ph.SR
|
http://creativecommons.org/licenses/by/4.0/
|
We present a joint analysis of TESS photometry and radial velocity measurements obtained from the Minerva-Australis facility for two short-period eclipsing binaries, TIC 48227288 and TIC 339607421. TIC 339607421 hosts an M-dwarf companion ($M_B = 0.294 \pm 0.013 \: M_{\odot}$, $R_B = 0.291 \pm 0.006 \: R_{\odot}$) orbiting an F6V star ($M_A=1.09 \pm 0.04 \: M_{\odot}$, $R_A=1.21^{+0.03}_{-0.02} \: R_{\odot}$). While TIC 48227288 contains a late K class companion ($M_B=0.635 \pm 0.037 \: M_{\odot}$, $R_B = 0.605 \pm 0.011 \: R_{\odot}$) orbiting an F3V star ($M_A = 1.36^{+0.06}_{-0.08} \: M_{\odot}$, $R_A = 1.61 \pm 0.03 \: R_{\odot}$). Both companions follow short period, near-circular orbits ($P_B = 2.4-3.0$ d, $e \approx 0.001$). Sky-projected obliquities for each system were derived using a classical analysis of the RV perturbation and the Reloaded Rossiter-McLaughlin (RRM) technique. The classical method indicates minor spin-orbit misalignment for both systems ($\lambda_A = -14.7^{+5.4}_{-5.9}$ deg and $-17.8^{+1.9}_{-2.0}$ deg for TIC 339607421 and TIC 48227288, respectively). The RRM analysis yields smaller obliquities ($\lambda_A = -8.2 \pm 0.2$ deg and $-9.5 \pm 0.2$ deg respectively), but confirms the minor misalignment inferred from the classical analysis. The findings of misaligned, circular orbits are notable even though the misalignments are not large, and suggest potential gaps in current models of binary formation and orbital evolution. As such, further investigation of these and similar systems appears warranted.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:25:13 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Wells",
"Tony",
""
],
[
"Addison",
"B. C.",
""
],
[
"Wittenmyer",
"R. A.",
""
],
[
"Wright",
"Duncan J.",
""
],
[
"Fairnington",
"Tyler R.",
""
],
[
"Dittmann",
"Jason",
""
],
[
"Horner",
"Jonathan",
""
],
[
"Kane",
"Stephen R.",
""
],
[
"Kielkopf",
"John",
""
],
[
"Plavchan",
"Peter",
""
],
[
"Shporer",
"Avi",
""
]
] |
2509.03827
|
Serge Stinckwich
|
Pierre Le Coz, Jia An Liu, Debarun Bhattacharjya, Georgina Curto, Serge Stinckwich
|
What Would an LLM Do? Evaluating Policymaking Capabilities of Large Language Models
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Large language models (LLMs) are increasingly being adopted in high-stakes domains. Their capacity to process vast amounts of unstructured data, explore flexible scenarios, and handle a diversity of contextual factors can make them uniquely suited to provide new insights for the complexity of social policymaking. This article evaluates whether LLMs' are aligned with domain experts (and among themselves) to inform social policymaking on the subject of homelessness alleviation - a challenge affecting over 150 million people worldwide. We develop a novel benchmark comprised of decision scenarios with policy choices across four geographies (South Bend, USA; Barcelona, Spain; Johannesburg, South Africa; Macau SAR, China). The policies in scope are grounded in the conceptual framework of the Capability Approach for human development. We also present an automated pipeline that connects the benchmarked policies to an agent-based model, and we explore the social impact of the recommended policies through simulated social scenarios. The paper results reveal promising potential to leverage LLMs for social policy making. If responsible guardrails and contextual calibrations are introduced in collaboration with local domain experts, LLMs can provide humans with valuable insights, in the form of alternative policies at scale.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:28:58 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Coz",
"Pierre Le",
""
],
[
"Liu",
"Jia An",
""
],
[
"Bhattacharjya",
"Debarun",
""
],
[
"Curto",
"Georgina",
""
],
[
"Stinckwich",
"Serge",
""
]
] |
2509.03828
|
Jaerong Ahn
|
Jaerong Ahn, Andrew Wen, Nan Wang, Heling Jia, Zhiyi Yue, Sunyang Fu and Hongfang Liu
|
An Agentic Model Context Protocol Framework for Medical Concept Standardization
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
The Observational Medical Outcomes Partnership (OMOP) common data model (CDM) provides a standardized representation of heterogeneous health data to support large-scale, multi-institutional research. One critical step in data standardization using OMOP CDM is the mapping of source medical terms to OMOP standard concepts, a procedure that is resource-intensive and error-prone. While large language models (LLMs) have the potential to facilitate this process, their tendency toward hallucination makes them unsuitable for clinical deployment without training and expert validation. Here, we developed a zero-training, hallucination-preventive mapping system based on the Model Context Protocol (MCP), a standardized and secure framework allowing LLMs to interact with external resources and tools. The system enables explainable mapping and significantly improves efficiency and accuracy with minimal effort. It provides real-time vocabulary lookups and structured reasoning outputs suitable for immediate use in both exploratory and production environments.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:32:22 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Ahn",
"Jaerong",
""
],
[
"Wen",
"Andrew",
""
],
[
"Wang",
"Nan",
""
],
[
"Jia",
"Heling",
""
],
[
"Yue",
"Zhiyi",
""
],
[
"Fu",
"Sunyang",
""
],
[
"Liu",
"Hongfang",
""
]
] |
2509.03829
|
Huhong Xian
|
Huhong Xian, Rui Liu, Berrak Sisman and Haizhou Li
|
NE-PADD: Leveraging Named Entity Knowledge for Robust Partial Audio Deepfake Detection via Attention Aggregation
| null | null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Different from traditional sentence-level audio deepfake detection (ADD), partial audio deepfake detection (PADD) requires frame-level positioning of the location of fake speech. While some progress has been made in this area, leveraging semantic information from audio, especially named entities, remains an underexplored aspect. To this end, we propose NE-PADD, a novel method for Partial Audio Deepfake Detection (PADD) that leverages named entity knowledge through two parallel branches: Speech Name Entity Recognition (SpeechNER) and PADD. The approach incorporates two attention aggregation mechanisms: Attention Fusion (AF) for combining attention weights and Attention Transfer (AT) for guiding PADD with named entity semantics using an auxiliary loss. Built on the PartialSpoof-NER dataset, experiments show our method outperforms existing baselines, proving the effectiveness of integrating named entity knowledge in PADD. The code is available at https://github.com/AI-S2-Lab/NE-PADD.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:33:00 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Xian",
"Huhong",
""
],
[
"Liu",
"Rui",
""
],
[
"Sisman",
"Berrak",
""
],
[
"Li",
"Haizhou",
""
]
] |
2509.03830
|
Kaizhen Tan
|
Kaizhen Tan, Yufan Wu, Yuxuan Liu, Haoran Zeng
|
A Multidimensional AI-powered Framework for Analyzing Tourist Perception in Historic Urban Quarters: A Case Study in Shanghai
| null | null | null | null |
cs.AI cs.CV cs.CY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Historic urban quarters play a vital role in preserving cultural heritage while serving as vibrant spaces for tourism and everyday life. Understanding how tourists perceive these environments is essential for sustainable, human-centered urban planning. This study proposes a multidimensional AI-powered framework for analyzing tourist perception in historic urban quarters using multimodal data from social media. Applied to twelve historic quarters in central Shanghai, the framework integrates focal point extraction, color theme analysis, and sentiment mining. Visual focus areas are identified from tourist-shared photos using a fine-tuned semantic segmentation model. To assess aesthetic preferences, dominant colors are extracted using a clustering method, and their spatial distribution across quarters is analyzed. Color themes are further compared between social media photos and real-world street views, revealing notable shifts. This divergence highlights potential gaps between visual expectations and the built environment, reflecting both stylistic preferences and perceptual bias. Tourist reviews are evaluated through a hybrid sentiment analysis approach combining a rule-based method and a multi-task BERT model. Satisfaction is assessed across four dimensions: tourist activities, built environment, service facilities, and business formats. The results reveal spatial variations in aesthetic appeal and emotional response. Rather than focusing on a single technical innovation, this framework offers an integrated, data-driven approach to decoding tourist perception and contributes to informed decision-making in tourism, heritage conservation, and the design of aesthetically engaging public spaces.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:35:14 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Tan",
"Kaizhen",
""
],
[
"Wu",
"Yufan",
""
],
[
"Liu",
"Yuxuan",
""
],
[
"Zeng",
"Haoran",
""
]
] |
2509.03831
|
Yilen G\'omez Maqueo Chew
|
Y. G\'omez Maqueo Chew, L. Hebb, H.C. Stempels, F.M. Walter, D.J. James, G.A. Feiden, R. Petrucci, T. Lister, I. Baraffe, M. Brodheim, F. Faedi, D.R. Anderson, R.A. Street, C. Hellier, K.G. Stassun
|
Discovery of the pre-main-sequence eclipsing binary MML 48
|
13 pages, 11 figures, accepted for publication in A&A
| null |
10.1051/0004-6361/202555622
| null |
astro-ph.SR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present the discovery of the eclipsing binary MML 48, which is a member of Upper Centaurus Lupus, has an associated age of 16 Myr, and is composed of two young, low-mass stars. We used space- and ground-based observations to characterize the system with both time-series photometry and spectroscopy. Given the extreme mass ratio between the stars, q_EB = 0.209 +- 0.014, we modeled a single-lined spectroscopic and eclipsing binary system. The orbital period, 2.0171068 +- 0.0000004 d, is measured from the highest precision light curves. We derive a primary mass of 1.2 +- 0.07 Msun using stellar models, and with radial velocities we measured a secondary mass of 0.2509 +- 0.0078 Msun. The radii are large, as expected for pre-main-sequence stars, and are measured as 1.574 +- 0.026 +- 0.050 Rsun and 0.587 +- 0.0095 +- 0.050 Rsun, for the primary and secondary stars, respectively. MML 48 joins the short list of known low-mass, pre-main-sequence eclipsing binaries (EBs), being one of only five systems with intermediate ages (15-25 Myr), and the system with the most extreme mass ratio. The primary star is currently at the "fusion bump", undergoing an over-production of energy in the core due to the build-up of 3He before reaching its equilibrium abundance set by the proton-proton (p-p) I chain. MML 48 A is the first young star in an eclipsing system that has been found during its fusion bump. MML 48 is thus an important benchmark for low-mass stellar evolution at a time when the stars are rapidly changing, which allows for a tight constraint on the corresponding isochrone given the uneven mass ratio.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:41:28 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chew",
"Y. Gómez Maqueo",
""
],
[
"Hebb",
"L.",
""
],
[
"Stempels",
"H. C.",
""
],
[
"Walter",
"F. M.",
""
],
[
"James",
"D. J.",
""
],
[
"Feiden",
"G. A.",
""
],
[
"Petrucci",
"R.",
""
],
[
"Lister",
"T.",
""
],
[
"Baraffe",
"I.",
""
],
[
"Brodheim",
"M.",
""
],
[
"Faedi",
"F.",
""
],
[
"Anderson",
"D. R.",
""
],
[
"Street",
"R. A.",
""
],
[
"Hellier",
"C.",
""
],
[
"Stassun",
"K. G.",
""
]
] |
2509.03832
|
Georgiy Lapin
|
Joseph Jackson, Georgiy Lapin, Jeremy E. Thompson
|
Gravity Well Echo Chamber Modeling With An LLM-Based Confirmation Bias Model
| null | null | null | null |
cs.SI cs.AI cs.CY
|
http://creativecommons.org/licenses/by/4.0/
|
Social media echo chambers play a central role in the spread of misinformation, yet existing models often overlook the influence of individual confirmation bias. An existing model of echo chambers is the "gravity well" model, which creates an analog between echo chambers and spatial gravity wells. We extend this established model by introducing a dynamic confirmation bias variable that adjusts the strength of pull based on a user's susceptibility to belief-reinforcing content. This variable is calculated for each user through comparisons between their posting history and their responses to posts of a wide range of viewpoints.
Incorporating this factor produces a confirmation-bias-integrated gravity well model that more accurately identifies echo chambers and reveals community-level markers of information health. We validated the approach on nineteen Reddit communities, demonstrating improved detection of echo chambers.
Our contribution is a framework for systematically capturing the role of confirmation bias in online group dynamics, enabling more effective identification of echo chambers. By flagging these high-risk environments, the model supports efforts to curb the spread of misinformation at its most common points of amplification.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:41:35 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Jackson",
"Joseph",
""
],
[
"Lapin",
"Georgiy",
""
],
[
"Thompson",
"Jeremy E.",
""
]
] |
2509.03833
|
Minghan Chu
|
Minghan Chu, Weicheng Qian
|
Towards Physics Constrained Deep Learning Based Turbulence Model Uncertainty Quantification
| null | null | null | null |
physics.flu-dyn
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Turbulence Models represent the workhorse for simulations used in engineering design and analysis. Despite their low computational cost and robustness, these models suffer from substantial predictive uncertainty, most of which is epistemic. At present, the Eigenspace Perturbation Method (EPM) is the only approach to estimate these turbulence model uncertainties, using physics based perturbation to the predicted Reynolds stresses. While the EPM address the question of how to perturb the Reynolds stresses for uncertainty estimation, it does not address how much to perturb. This shortcoming leads to very generous uncertainty bounds that result in sub-optimal designs. In this investigation, we use Convolutional Neural Networks (CNN) to predict the discrepancy between predicted and actual turbulent flows. These can be utilized to modulate the degree of the perturbations in the EPM leading to a Physics Constrained Deep Learning approach for Reynolds Averaged Navier Stokes model uncertainty quantification. We test this approach on turbulent flows over aero-foils and periodic hills to show the efficacy of our approach.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:47:46 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chu",
"Minghan",
""
],
[
"Qian",
"Weicheng",
""
]
] |
2509.03834
|
Daniel Menasche
|
Lucas Lopes Felipe and Konstantin Avrachenkov and Daniel Sadoc Menasche
|
From Leiden to Pleasure Island: The Constant Potts Model for Community Detection as a Hedonic Game
|
Manuscript submitted to Physica A: Statistical Mechanics and its Applications
| null | null | null |
cs.LG cs.AI cs.GT
|
http://creativecommons.org/licenses/by/4.0/
|
Community detection is one of the fundamental problems in data science which consists of partitioning nodes into disjoint communities. We present a game-theoretic perspective on the Constant Potts Model (CPM) for partitioning networks into disjoint communities, emphasizing its efficiency, robustness, and accuracy. Efficiency: We reinterpret CPM as a potential hedonic game by decomposing its global Hamiltonian into local utility functions, where the local utility gain of each agent matches the corresponding increase in global utility. Leveraging this equivalence, we prove that local optimization of the CPM objective via better-response dynamics converges in pseudo-polynomial time to an equilibrium partition. Robustness: We introduce and relate two stability criteria: a strict criterion based on a novel notion of robustness, requiring nodes to simultaneously maximize neighbors and minimize non-neighbors within communities, and a relaxed utility function based on a weighted sum of these objectives, controlled by a resolution parameter. Accuracy: In community tracking scenarios, where initial partitions are used to bootstrap the Leiden algorithm with partial ground-truth information, our experiments reveal that robust partitions yield higher accuracy in recovering ground-truth communities.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:50:54 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Felipe",
"Lucas Lopes",
""
],
[
"Avrachenkov",
"Konstantin",
""
],
[
"Menasche",
"Daniel Sadoc",
""
]
] |
2509.03835
|
Wei Zhang
|
Wei Zhang, Zhenming Yu, Hongyu Huang, Xiangyong Dong, Yan Ma, Yongli Zhao, Shanguo Huang, and Kun Xu
|
Robust End-to-End FSO Transmission with Joint Coding Modulation and BiLSTM-Based Channel Modeling under Atmospheric Turbulence
| null | null | null | null |
physics.optics
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Free space optical (FSO) communication is considered a promising solution in next_generation communication networks. However, its performance is significantly influenced by atmospheric turbulence. To enhance system robustness to turbulence, we propose a turbulence_robust end_to_end FSO communication system (TRFSO) that integrates a data-driven channel model with joint source_channel coding modulation (JSCCM). Specifically, a bidirectional long short-term memory (BiLSTM)_based channel model is developed and trained on data collected over a physical FSO link under varying turbulence conditions. This model accurately captures real_world channel distortions, achieving a minimum Kullback_Leibler (KL) divergence of 0.0019 in amplitude distribution matching. Experimental results show that the TRFSO system trained with the BiLSTM_based channel model outperforms the same architecture trained under the additive white Gaussian noise (AWGN) channel, achieving an average 3.5 dB improvement in multi-scale structural similarity (MS_SSIM) under strong atmospheric turbulence. These results demonstrate the effectiveness of the proposed TRFSO in achieving robust and reliable transmission under dynamic atmospheric turbulence.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:56:31 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhang",
"Wei",
""
],
[
"Yu",
"Zhenming",
""
],
[
"Huang",
"Hongyu",
""
],
[
"Dong",
"Xiangyong",
""
],
[
"Ma",
"Yan",
""
],
[
"Zhao",
"Yongli",
""
],
[
"Huang",
"Shanguo",
""
],
[
"Xu",
"Kun",
""
]
] |
2509.03836
|
Bingxin Zhang
|
Bingxin Zhang, Han Zhang, Kun Yang, Yizhe Zhao, and Kezhi Wang
|
On the Performance Analysis of Pinching-Antenna-Enabled SWIPT Systems
| null | null | null | null |
eess.SY cs.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we studies the performance of a novel simultaneous wireless information and power transfer (SWIPT) system enabled by a flexible pinching-antenna. To support flexible deployment and optimize energy-rate performance, we propose three practical pinching antenna placement-schemes: the edge deployment scheme (EDS), the center deployment scheme (CDS), and the diagonal deployment scheme (DDS). Moreover, a hybrid time-switching (TS) and power-splitting (PS) protocol is introduced, allowing dynamic adjustment between energy harvesting and information decoding. Under each deployment strategy and the transmission protocol, closed-form expressions for the average harvested energy and average achievable rate of a randomly located user equipment (UE) are derived based on the optimal positioning of the pinching-antenna. Numerical simulations confirm the accuracy of the theoretical analysis and illustrate the trade-off between rate and energy harvesting under different schemes.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:56:43 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhang",
"Bingxin",
""
],
[
"Zhang",
"Han",
""
],
[
"Yang",
"Kun",
""
],
[
"Zhao",
"Yizhe",
""
],
[
"Wang",
"Kezhi",
""
]
] |
2509.03837
|
Kimia Ehsani
|
Kimia Ehsani, Walid Saad
|
Vehicle-to-Infrastructure Collaborative Spatial Perception via Multimodal Large Language Models
|
Accepted at IEEE GLOBECOM 2025
| null | null | null |
cs.LG cs.IT math.IT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Accurate prediction of communication link quality metrics is essential for vehicle-to-infrastructure (V2I) systems, enabling smooth handovers, efficient beam management, and reliable low-latency communication. The increasing availability of sensor data from modern vehicles motivates the use of multimodal large language models (MLLMs) because of their adaptability across tasks and reasoning capabilities. However, MLLMs inherently lack three-dimensional spatial understanding. To overcome this limitation, a lightweight, plug-and-play bird's-eye view (BEV) injection connector is proposed. In this framework, a BEV of the environment is constructed by collecting sensing data from neighboring vehicles. This BEV representation is then fused with the ego vehicle's input to provide spatial context for the large language model. To support realistic multimodal learning, a co-simulation environment combining CARLA simulator and MATLAB-based ray tracing is developed to generate RGB, LiDAR, GPS, and wireless signal data across varied scenarios. Instructions and ground-truth responses are programmatically extracted from the ray-tracing outputs. Extensive experiments are conducted across three V2I link prediction tasks: line-of-sight (LoS) versus non-line-of-sight (NLoS) classification, link availability, and blockage prediction. Simulation results show that the proposed BEV injection framework consistently improved performance across all tasks. The results indicate that, compared to an ego-only baseline, the proposed approach improves the macro-average of the accuracy metrics by up to 13.9%. The results also show that this performance gain increases by up to 32.7% under challenging rainy and nighttime conditions, confirming the robustness of the framework in adverse settings.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:57:47 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Ehsani",
"Kimia",
""
],
[
"Saad",
"Walid",
""
]
] |
2509.03838
|
Julio Alberto M\'endez-Zavaleta
|
Mois\'es Bravo-Gaete, Luis Guajardo, Daniel F. Higuita-Borja, Julio A. M\'endez-Zavaleta
|
Transport coefficients of charged Gauss-Bonnet black holes with arbitrary topology
|
29 pages, 5 figures
| null | null | null |
hep-th gr-qc
|
http://creativecommons.org/licenses/by/4.0/
|
In this study, we present a novel family of exact black hole solutions constructed in the context of five-dimensional Gauss-Bonnet gravity. These solutions add a non-linear charge to the Ba\~{n}ados-Teitelboim-Zanelli-like configurations known to exist with arbitrary Thurston horizon geometry. We establish constraints on the parameter space defining physically viable black holes, aligning with the standard energy conditions. An explicit proof of the first law of thermodynamics within our scenario is provided. We also employ holographic techniques to characterize the DC conductivities for the distinct horizon geometries, identifying a critical temperature indicative of phase transitions and exploring pertinent limits.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:58:42 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Bravo-Gaete",
"Moisés",
""
],
[
"Guajardo",
"Luis",
""
],
[
"Higuita-Borja",
"Daniel F.",
""
],
[
"Méndez-Zavaleta",
"Julio A.",
""
]
] |
2509.03839
|
Daisuke Inoue
|
Daisuke Inoue, Tadayoshi Matsumori, Gouhei Tanaka, Yuji Ito
|
Reservoir Predictive Path Integral Control for Unknown Nonlinear Dynamics
|
Submitted to IEEE for possible publication, 13 pages, 7 figures
| null | null | null |
eess.SY cs.LG cs.SY math.OC nlin.CD
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Neural networks capable of approximating complex nonlinearities have found extensive application in data-driven control of nonlinear dynamical systems. However, fast online identification and control of unknown dynamics remain central challenges. This paper integrates echo-state networks (ESNs) -- reservoir computing models implemented with recurrent neural networks -- and model predictive path integral (MPPI) control -- sampling-based variants of model predictive control -- to meet these challenges. The proposed reservoir predictive path integral (RPPI) enables fast learning of nonlinear dynamics with ESN and exploits the learned nonlinearities directly in parallelized MPPI control computation without linearization approximations. The framework is further extended to uncertainty-aware RPPI (URPPI), which leverages ESN uncertainty to balance exploration and exploitation: exploratory inputs dominate during early learning, while exploitative inputs prevail as model confidence grows. Experiments on controlling the Duffing oscillator and four-tank systems demonstrate that URPPI improves control performance, reducing control costs by up to 60% compared to traditional quadratic programming-based model predictive control methods.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:05:17 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Inoue",
"Daisuke",
""
],
[
"Matsumori",
"Tadayoshi",
""
],
[
"Tanaka",
"Gouhei",
""
],
[
"Ito",
"Yuji",
""
]
] |
2509.03840
|
Nour Alnajjarine
|
Nour Alnajjarine and Michel Lavrauw
|
Nets of conics containing a double line in $\mathrm{PG}(2,q)$, $q$ even
| null | null | null | null |
math.CO math.AG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we complete the classification of nets of conics containing at least one double line in $\mathrm{PG}(2,q)$ for $q$ even. This classification is obtained by classifying the orbits of planes in $\mathrm{PG}(5,q)$ that intersect the nucleus plane in at least one point, under the action of the group that stabilizes the Veronese surface in $\mathrm{PG}(5,q)$ for $q\geq 4$ even. As a result, we show the existence of $18$ nets containing at least one double line, $9$ of which have an empty base.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:06:04 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Alnajjarine",
"Nour",
""
],
[
"Lavrauw",
"Michel",
""
]
] |
2509.03841
|
Huichao Song
|
Shujun Zhao, Yiyang Peng, Ulrich W. Heinz, Huichao Song
|
Exploring the fluid behavior in p+p collisions at $\sqrt{s}=13 \mathrm{TeV}$ with viscous anisotropic hydrodynamics
|
6 pages,3 figures
| null | null | null |
nucl-th hep-ph nucl-ex
|
http://creativecommons.org/publicdomain/zero/1.0/
|
The applicability of hydrodynamics in small collision systems remains controversial due to the small size and short lifetime of the system. In this letter, we employ viscous anisotropic hydrodynamics (VAH), which incorporates large pressure anisotropies, to study the collectivity in p+p collisions at $\sqrt{s}=13 \mathrm{TeV}$.VAH provides a good description for $v_{2}\{2\}$ and $v_{3}\{2\}$ over a wide range of multiplicities and correctly reproduces the experimentally observed negative $c_{2}\{4\}$. Traditional second-order viscous hydrodynamics (VH), on the other hand, can describe the measurements, in particular the negative $c_{2}\{4\}$, only with model parameters for which the bulk of the evolution is characterized by large values of the shear Knudsen number. It also can not capture the large longitudinal/transverse pressure anisotropy during the early evolution. These demonstrate the failure of traditional viscous hydrodynamics in small collision systems and establish viscous anisotropic hydrodynamics as a more reliable framework to describe the bulk evolution and the observed anisotropic flow in p-p collisions at the LHC.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:07:16 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhao",
"Shujun",
""
],
[
"Peng",
"Yiyang",
""
],
[
"Heinz",
"Ulrich W.",
""
],
[
"Song",
"Huichao",
""
]
] |
2509.03842
|
Guanglu Jia Dr
|
Guanglu Jia, Ceng Zhang, Gregory S. Chirikjian
|
INGRID: Intelligent Generative Robotic Design Using Large Language Models
|
15 pages, 6 figures
| null | null | null |
cs.RO cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The integration of large language models (LLMs) into robotic systems has accelerated progress in embodied artificial intelligence, yet current approaches remain constrained by existing robotic architectures, particularly serial mechanisms. This hardware dependency fundamentally limits the scope of robotic intelligence. Here, we present INGRID (Intelligent Generative Robotic Design), a framework that enables the automated design of parallel robotic mechanisms through deep integration with reciprocal screw theory and kinematic synthesis methods. We decompose the design challenge into four progressive tasks: constraint analysis, kinematic joint generation, chain construction, and complete mechanism design. INGRID demonstrates the ability to generate novel parallel mechanisms with both fixed and variable mobility, discovering kinematic configurations not previously documented in the literature. We validate our approach through three case studies demonstrating how INGRID assists users in designing task-specific parallel robots based on desired mobility requirements. By bridging the gap between mechanism theory and machine learning, INGRID enables researchers without specialized robotics training to create custom parallel mechanisms, thereby decoupling advances in robotic intelligence from hardware constraints. This work establishes a foundation for mechanism intelligence, where AI systems actively design robotic hardware, potentially transforming the development of embodied AI systems.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:08:01 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Jia",
"Guanglu",
""
],
[
"Zhang",
"Ceng",
""
],
[
"Chirikjian",
"Gregory S.",
""
]
] |
2509.03843
|
Kane Sjoberg
|
Kane M. Sjoberg, Johnny Esteves, Christopher W. Stubbs
|
PANDORA: New avenues in sub-percent photometric precision for ground based astronomy
|
10 pages, 6 figures, presented at SPIE Optics + Photonics 2025
| null | null | null |
astro-ph.IM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Precision photometric calibration is key to a number of astrophysical research areas such as supernova cosmology and dark energy studies. In the age of large surveys, pushing the limits of photometry is a worthy challenge, with traditional celestial calibrators limited to approximately percent precision. Here we present PANDORA, a novel calibration source that achieves sub-percent photometric precision across the optical spectrum with a selectable dynamic range and high optical efficiency. Building on proven collimated beam projector designs for relative throughput calibration of survey telescopes, PANDORA is capable of delivering photons anywhere in the 350-1100 nm spectral range. It uses selectable neutral density filters and sequential NIST-calibrated photodiode feedback to internally self-calibrate prior to operation, and is illuminated by an Energetiq EQ-99X-FC broadband laser-driven light source. We present here an overview of the system design, operation, and laboratory characterization results. PANDORA will begin science operations this year to support the upcoming LS4 survey on the ESO 1m Schmidt telescope.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:09:30 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Sjoberg",
"Kane M.",
""
],
[
"Esteves",
"Johnny",
""
],
[
"Stubbs",
"Christopher W.",
""
]
] |
2509.03844
|
Fazal Badshah
|
Fazal Badshah, Imed Boukhris, Mohamed Sultan Al-Buriah, Yuan Zhou, Muhammad Idrees, Ziauddin
|
Tunneling induced giant Photonic Spin Hall effect in quantum wells
|
7 Pages, 6 Figures
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by-sa/4.0/
|
We propose a theoretical investigation of the photonic spin Hall effect (PSHE) in a mid-infrared probe field by employing an asymmetric double AlGaAsGaAs quantum well as the intracavity medium. The system is designed such that an external control beam together with tunable tunneling barriers regulates the quantum interference of the probe tunneling process. This configuration enables precise manipulation of the PSHE for both horizontally and vertically polarized components of light. Our analysis reveals the emergence of a giant horizontal PSHE in the quantum well based cavity system. Moreover, by incorporating absorptive and gain-assisted cavity slabs, the horizontal PSHE is further amplified, leading to an even more pronounced photonic spin separation. The results provide novel insights into light matter interactions in semiconductor quantum wells and suggest an effective route for enhancing and controlling the PSHE in mid-infrared photonic devices.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:13:00 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Badshah",
"Fazal",
""
],
[
"Boukhris",
"Imed",
""
],
[
"Al-Buriah",
"Mohamed Sultan",
""
],
[
"Zhou",
"Yuan",
""
],
[
"Idrees",
"Muhammad",
""
],
[
"Ziauddin",
"",
""
]
] |
2509.03845
|
Yang Chen
|
Yang Chen, Xiao Lin, Bo Yan, Libo Zhang, Jiamou Liu, Neset \"Ozkan Tan, Michael Witbrock
|
Meta-Inverse Reinforcement Learning for Mean Field Games via Probabilistic Context Variables
|
Accepted to AAAI 2024
| null | null | null |
cs.LG cs.AI cs.GT
|
http://creativecommons.org/licenses/by/4.0/
|
Designing suitable reward functions for numerous interacting intelligent agents is challenging in real-world applications. Inverse reinforcement learning (IRL) in mean field games (MFGs) offers a practical framework to infer reward functions from expert demonstrations. While promising, the assumption of agent homogeneity limits the capability of existing methods to handle demonstrations with heterogeneous and unknown objectives, which are common in practice. To this end, we propose a deep latent variable MFG model and an associated IRL method. Critically, our method can infer rewards from different yet structurally similar tasks without prior knowledge about underlying contexts or modifying the MFG model itself. Our experiments, conducted on simulated scenarios and a real-world spatial taxi-ride pricing problem, demonstrate the superiority of our approach over state-of-the-art IRL methods in MFGs.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:13:11 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chen",
"Yang",
""
],
[
"Lin",
"Xiao",
""
],
[
"Yan",
"Bo",
""
],
[
"Zhang",
"Libo",
""
],
[
"Liu",
"Jiamou",
""
],
[
"Tan",
"Neset Özkan",
""
],
[
"Witbrock",
"Michael",
""
]
] |
2509.03846
|
Rownak Chowdhury
|
Md Rownak Hossain Chowdhury, Mostafizur Rahman
|
Hardware-Aware Data and Instruction Mapping for AI Tasks: Balancing Parallelism, I/O and Memory Tradeoffs
| null | null | null | null |
cs.AR cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We introduce a mapping framework for deep learning inference that takes advantage of predictable neural network behavior to plan both computation and communication ahead of time. The framework generates a unified stream of instructions and data, enabling the hardware to execute operations and route information on its own, without frequent involvement from the host and with minimal off-chip memory use. This naturally reduces reliance on I/O, off-chip memory, and host control. By leveraging fine-grained message passing on a programmable, message-based compute architecture, the framework keeps data movement local and coordinates computation across the array using techniques such as stationary-weight reuse, in-array multicasting, and staged reductions. Applied to VGG-19, the framework sustains high utilization (88 to 92 percent), with over 97 percent of messages generated internally and nearly 89 percent of time consumed on-chip transfers. Computation throughput scales beyond 1 TFLOP/s on larger arrays, while traffic reductions from reuse and local aggregation reach up to 100 MB per layer. Overall, the results highlight the effectiveness of streaming-based computation and show how our mapper enables this execution style by tightly coordinating data and instruction flow across the hardware.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:14:16 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chowdhury",
"Md Rownak Hossain",
""
],
[
"Rahman",
"Mostafizur",
""
]
] |
2509.03847
|
Trong Hoang Do
|
Do Trong Hoang, Vadim E. Levit, Eugen Mandrescu
|
Structural properties and characterizations of $\mathbf{W}_{p}$ class
|
13 pages
| null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
We establish new characterizations of graphs belonging to the $\mathbf{W}_p$ class. In addition, we characterize locally triangle-free $\alpha$-critical graphs in this class. As a consequence, our results yield a partial answer to a question raised by Plummer \cite{Plum93} in the case $p=2$.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:15:36 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Hoang",
"Do Trong",
""
],
[
"Levit",
"Vadim E.",
""
],
[
"Mandrescu",
"Eugen",
""
]
] |
2509.03848
|
Rodrigo Oliveira Zacarias
|
Rodrigo Oliveira Zacarias, Rodrigo Pereira dos Santos, Patricia Lago
|
Towards an Understanding of Developer Experience-Driven Transparency in Software Ecosystems
|
36 pages Submitted to the ACM Transactions on Software Engineering and Methodology. 2025
| null | null | null |
cs.SE cs.HC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Software ecosystems (SECO) have become a dominant paradigm in the software industry, enabling third-party developers to co-create value through complementary components and services. While Developer Experience (DX) is increasingly recognized as critical for sustainable SECO, transparency remains an underexplored factor shaping how developers perceive and interact with ecosystems. Existing studies acknowledge transparency as essential for trust, fairness, and engagement, yet its relationship with DX has not been systematically conceptualized. Hence, this work aims to advance the understanding of transparency in SECO from a developer-centered perspective. To this end, we propose SECO-TransDX (Transparency in Software Ecosystems from a Developer Experience Perspective), a conceptual model that introduces the notion of DX-driven transparency. The model identifies 63 interrelated concepts, including conditioning factors, ecosystem procedures, artifacts, and relational dynamics that influence how transparency is perceived and constructed during developer interactions. SECO-TransDX was built upon prior research and refined through a Delphi study with experts from academia and industry. It offers a structured lens to examine how transparency mediates DX across technical, social, and organizational layers. For researchers, it lays the groundwork for future studies and tool development; for practitioners, it supports the design of trustworthy, developer-centered platforms that improve transparency and foster long-term engagement in SECO.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:17:22 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zacarias",
"Rodrigo Oliveira",
""
],
[
"Santos",
"Rodrigo Pereira dos",
""
],
[
"Lago",
"Patricia",
""
]
] |
2509.03849
|
Boyang Liu
|
Bo Chen, Xiaojing Jin, Jiatao Tan, and Boyang Liu
|
Merons and Meroniums in Spin-Orbit Coupled Bose Gases
|
5 pages, 4 figures
| null | null | null |
cond-mat.quant-gas
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this work we demonstrate the existence of new types of topological states in a two-component Bose gas with Rashba spin-orbit couplings. We find four types of topological structures exist in this system. First, the half vortex (HV), which have been found and discussed in similar systems. Second, the spherical wave half vortex (SWHV), which is different from the HV by a factor of $e^{ikr}$ in the wave function. The spin configurations show that HV and SWHV are both merons. Of particular interests are the third and fourth types, which are called double peak (DP) and spin spiral (SS) phases. They are both combinations of meron and antimeron, and hence named as meroniums. These two phases demonstrate intriguing spin density patterns. Finally, the phase diagram is obtained and the stability of the meronium state is discussed.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:23:12 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Chen",
"Bo",
""
],
[
"Jin",
"Xiaojing",
""
],
[
"Tan",
"Jiatao",
""
],
[
"Liu",
"Boyang",
""
]
] |
2509.03850
|
Justin Kur
|
Justin Kur, Kaiqi Zhao
|
Data-Augmented Quantization-Aware Knowledge Distillation
|
10 pages, 2 figures
| null | null | null |
cs.LG cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Quantization-aware training (QAT) and Knowledge Distillation (KD) are combined to achieve competitive performance in creating low-bit deep learning models. Existing KD and QAT works focus on improving the accuracy of quantized models from the network output perspective by designing better KD loss functions or optimizing QAT's forward and backward propagation. However, limited attention has been given to understanding the impact of input transformations, such as data augmentation (DA). The relationship between quantization-aware KD and DA remains unexplored. In this paper, we address the question: how to select a good DA in quantization-aware KD, especially for the models with low precisions? We propose a novel metric which evaluates DAs according to their capacity to maximize the Contextual Mutual Information--the information not directly related to an image's label--while also ensuring the predictions for each class are close to the ground truth labels on average. The proposed method automatically ranks and selects DAs, requiring minimal training overhead, and it is compatible with any KD or QAT algorithm. Extensive evaluations demonstrate that selecting DA strategies using our metric significantly improves state-of-the-art QAT and KD works across various model architectures and datasets.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:24:35 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Kur",
"Justin",
""
],
[
"Zhao",
"Kaiqi",
""
]
] |
2509.03851
|
Doudou Hei
|
Doudou Hei, Xinmin Hou, Yue Ma
|
The generalized Tur\'an number for K_3 in graphs without suspensions of a path on five vertices
| null | null | null | null |
math.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Given graphs $H$ and $F$, the generalized Tur\'an number $\ex(n, H, F)$ is defined as
the maximum number of copies of $H$ in an $n$-vertex graph that contains no copy of $F$.
The suspension $\widehat{F}$ of a graph $F$ is
obtained by adding a new vertex that is adjacent to every vertex of $F$.
Mubayi and Mukherjee (2023, DM) conjectured that $\ex(n, K_3, \widehat{P_k})=\left\lfloor \frac{k-2}{2}\right\rfloor \cdot \frac{n^2}{8}+o(n^2)$, where $P_k$ is a
path on $k\ge 4$ vertices. Using the triangle removal
lemma, they verified this conjecture for $k=4,5,6$.
Later, Mukherjee (2024, DM) established the exact value $\ex(n, K_3, \widehat{P_4})=\left\lfloor n^2/8\right\rfloor$.
In this paper, using the stability method, we determine the exact value of $\ex(n, K_3, \widehat{P_5})$ by showing that for sufficiently large $n$, $\ex(n,K_3, \widehat{P_5})=\left\lfloor n^2/8\right\rfloor.$
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:26:40 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Hei",
"Doudou",
""
],
[
"Hou",
"Xinmin",
""
],
[
"Ma",
"Yue",
""
]
] |
2509.03852
|
Ling Chen
|
Binqing Wu, Zongjiang Shang, Jianlong Huang, Ling Chen
|
MillGNN: Learning Multi-Scale Lead-Lag Dependencies for Multi-Variate Time Series Forecasting
|
Accepted by CIKM 2025
| null | null | null |
cs.LG cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Multi-variate time series (MTS) forecasting is crucial for various applications. Existing methods have shown promising results owing to their strong ability to capture intra- and inter-variate dependencies. However, these methods often overlook lead-lag dependencies at multiple grouping scales, failing to capture hierarchical lead-lag effects in complex systems. To this end, we propose MillGNN, a novel \underline{g}raph \underline{n}eural \underline{n}etwork-based method that learns \underline{m}ult\underline{i}ple grouping scale \underline{l}ead-\underline{l}ag dependencies for MTS forecasting, which can comprehensively capture lead-lag effects considering variate-wise and group-wise dynamics and decays. Specifically, MillGNN introduces two key innovations: (1) a scale-specific lead-lag graph learning module that integrates cross-correlation coefficients and dynamic decaying features derived from real-time inputs and time lags to learn lead-lag dependencies for each scale, which can model evolving lead-lag dependencies with statistical interpretability and data-driven flexibility; (2) a hierarchical lead-lag message passing module that passes lead-lag messages at multiple grouping scales in a structured way to simultaneously propagate intra- and inter-scale lead-lag effects, which can capture multi-scale lead-lag effects with a balance of comprehensiveness and efficiency. Experimental results on 11 datasets demonstrate the superiority of MillGNN for long-term and short-term MTS forecasting, compared with 16 state-of-the-art methods.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:28:42 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Wu",
"Binqing",
""
],
[
"Shang",
"Zongjiang",
""
],
[
"Huang",
"Jianlong",
""
],
[
"Chen",
"Ling",
""
]
] |
2509.03853
|
Yuexi Wang
|
Haoyu Jiang, Yuexi Wang, Yun Yang
|
Simulation-based Inference via Langevin Dynamics with Score Matching
| null | null | null | null |
stat.ME stat.CO stat.ML
|
http://creativecommons.org/licenses/by/4.0/
|
Simulation-based inference (SBI) enables Bayesian analysis when the likelihood is intractable but model simulations are available. Recent advances in statistics and machine learning, including Approximate Bayesian Computation and deep generative models, have expanded the applicability of SBI, yet these methods often face challenges in moderate to high-dimensional parameter spaces. Motivated by the success of gradient-based Monte Carlo methods in Bayesian sampling, we propose a novel SBI method that integrates score matching with Langevin dynamics to explore complex posterior landscapes more efficiently in such settings. Our approach introduces tailored score-matching procedures for SBI, including a localization scheme that reduces simulation costs and an architectural regularization that embeds the statistical structure of log-likelihood scores to improve score-matching accuracy. We provide theoretical analysis of the method and illustrate its practical benefits on benchmark tasks and on more challenging problems in moderate to high dimensions, where it performs favorably compared to existing approaches.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:28:50 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Jiang",
"Haoyu",
""
],
[
"Wang",
"Yuexi",
""
],
[
"Yang",
"Yun",
""
]
] |
2509.03854
|
Joshua J Sanchez
|
Joshua J. Sanchez and Raagya Aurora and Daniel Bennett and Daniel T. Larson and Efthimios Kaxiras and Riccardo Comin
|
Switching topological states via uniaxial strain in 2D materials
|
6 pages, 4 figures
| null | null | null |
cond-mat.str-el
|
http://creativecommons.org/licenses/by/4.0/
|
In topological materials, dissipationless edge currents are protected against local defect scattering by the bulk inverted band structure and band gap. We propose that large uniaxial strain can effectively switch a 2D Chern insulator to a topologically trivial state. Further, we suggest that the boundary between strained and unstrained regions of a sample can act as a new edge for dissipationless current flow. Using density functional theory (DFT) calculations we demonstrate the strain-tunability of the monolayer MnBi2S2T2 band structure and the switching of the Chern number. We combine uniaxial and biaxial strain results to map out the strain-tuned topological phase diagram.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:29:22 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Sanchez",
"Joshua J.",
""
],
[
"Aurora",
"Raagya",
""
],
[
"Bennett",
"Daniel",
""
],
[
"Larson",
"Daniel T.",
""
],
[
"Kaxiras",
"Efthimios",
""
],
[
"Comin",
"Riccardo",
""
]
] |
2509.03855
|
Zhouyi Zhou
|
Zhouyi Zhou, Zhili Liu, Shancong Zhang, Jiemin Li, Dengke Du, Mengke Sun, Zhiqiang Wang, Hongyan Liu, Guogai Xu
|
Towards Deterministic Sub-0.5 us Response on Linux through Interrupt Isolation
|
9 pages, 11 figures
| null | null | null |
cs.OS
|
http://creativecommons.org/licenses/by/4.0/
|
Real-time responsiveness in Linux is often constrained by interrupt contention and timer handling overhead, making it challenging to achieve sub-microsecond latency. This work introduces an interrupt isolation approach that centralizes and minimizes timer interrupt interference across CPU cores. By enabling a dedicated API to selectively invoke timer handling routines and suppress non-critical inter-processor interrupts, our design significantly reduces jitter and response latency. Experiments conducted on an ARM-based multicore platform demonstrate that the proposed mechanism consistently achieves sub-0.5 us response times, outperforming conventional Linux PREEMPT-RT configurations. These results highlight the potential of interrupt isolation as a lightweight and effective strategy for deterministic real-time workloads in general-purpose operating systems.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:31:31 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Zhou",
"Zhouyi",
""
],
[
"Liu",
"Zhili",
""
],
[
"Zhang",
"Shancong",
""
],
[
"Li",
"Jiemin",
""
],
[
"Du",
"Dengke",
""
],
[
"Sun",
"Mengke",
""
],
[
"Wang",
"Zhiqiang",
""
],
[
"Liu",
"Hongyan",
""
],
[
"Xu",
"Guogai",
""
]
] |
2509.03856
|
Pei-Zi Zhao
|
X. Y. Sun, P. Z. Zhao
|
Decoherence mitigation for geometric quantum computation
|
8 pages, 2 figures
|
Phys. Rev. A 112, 032403 (2025)
|
10.1103/7y8p-85k3
| null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Geometric phases depend only on the evolution path determined by the closed circuit in the projective Hilbert space but not on evolution details of the quantum system, leading to geometric quantum computation possessing some intrinsic robustness against control errors. Coordinated with dynamical decoupling, geometric quantum computation admits additional resilience to the environment-induced decoherence. However, the previous schemes of geometric quantum computation protected by dynamical decoupling require multiple physical qubits to encode a logical qubit, which undoubtedly increases the consumption of physical-qubit resources and the difficulty in the implementation of the logical-qubit manipulation based on physical-qubit driving. In this work, we put forward a scheme of decoherence-mitigated geometric quantum computation based only on physical qubits rather than logical qubits, hence avoiding the additional overhead of physical-qubit resources for logical-qubit encoding as well as the difficulty in the manipulation of logical qubits. Moreover, our scheme focuses on the most general interaction between an individual qubit and its environment so that it mitigates not just dephasing noise but rather regular decoherence. Our proposal thus represents a more realistic and effective approach towards the realization of geometric control with decoherence mitigation.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:31:50 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Sun",
"X. Y.",
""
],
[
"Zhao",
"P. Z.",
""
]
] |
2509.03857
|
Syed Bahauddin Alam
|
Kishor Datta Gupta, Mohd Ariful Haque, Hasmot Ali, Marufa Kamal, Syed Bahauddin Alam, and Mohammad Ashiqur Rahman
|
Continuous Monitoring of Large-Scale Generative AI via Deterministic Knowledge Graph Structures
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Generative AI (GEN AI) models have revolutionized diverse application domains but present substantial challenges due to reliability concerns, including hallucinations, semantic drift, and inherent biases. These models typically operate as black-boxes, complicating transparent and objective evaluation. Current evaluation methods primarily depend on subjective human assessment, limiting scalability, transparency, and effectiveness. This research proposes a systematic methodology using deterministic and Large Language Model (LLM)-generated Knowledge Graphs (KGs) to continuously monitor and evaluate GEN AI reliability. We construct two parallel KGs: (i) a deterministic KG built using explicit rule-based methods, predefined ontologies, domain-specific dictionaries, and structured entity-relation extraction rules, and (ii) an LLM-generated KG dynamically derived from real-time textual data streams such as live news articles. Utilizing real-time news streams ensures authenticity, mitigates biases from repetitive training, and prevents adaptive LLMs from bypassing predefined benchmarks through feedback memorization. To quantify structural deviations and semantic discrepancies, we employ several established KG metrics, including Instantiated Class Ratio (ICR), Instantiated Property Ratio (IPR), and Class Instantiation (CI). An automated real-time monitoring framework continuously computes deviations between deterministic and LLM-generated KGs. By establishing dynamic anomaly thresholds based on historical structural metric distributions, our method proactively identifies and flags significant deviations, thus promptly detecting semantic anomalies or hallucinations. This structured, metric-driven comparison between deterministic and dynamically generated KGs delivers a robust and scalable evaluation framework.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:34:49 GMT"
}
] | 2025-09-05T00:00:00 |
[
[
"Gupta",
"Kishor Datta",
""
],
[
"Haque",
"Mohd Ariful",
""
],
[
"Ali",
"Hasmot",
""
],
[
"Kamal",
"Marufa",
""
],
[
"Alam",
"Syed Bahauddin",
""
],
[
"Rahman",
"Mohammad Ashiqur",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.