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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2508.11088
|
Eva Laplace
|
Eva Laplace and Vincent A. Bronner and Fabian R. N. Schneider and Philipp Podsiadlowski
|
Pulsations change the structures of massive stars before they explode: interpreting the nearby supernova SN 2023ixf
|
Submitted to AAS journals. This work is the follow-up of a companion paper on arXiv (Bronner, Laplace et al. 2025, arXiv:2508.11077)
| null | null | null |
astro-ph.HE astro-ph.SR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Massive red supergiants (RSGs) are known to become hydrodynamically unstable before they explode. Still, the vast majority of supernova (SN) models assume RSG progenitors in hydrostatic equilibrium. Here, we self-consistently follow the hydrodynamic evolution of RSGs with different masses and the development of radial envelope pulsations. Pulsations significantly alter the observable pre- and post-SN properties, and their importance increases substantially as a function of initial mass. We demonstrate that inferring core masses, let alone initial masses, from a single pre-SN luminosity and effective temperature of high-mass RSGs is inadvisable, as these can vary by an order of magnitude during the pulsation. We find that pulsations can naturally lead to ``early-excess" emission in SN light curves and to variations in early photospheric velocities, which can help break degeneracies in type-II SNe. We compare to SN~2023ixf, for which a pulsating RSG progenitor was discovered. We demonstrate that its pre- and post-SN characteristics agree very well with our exploding pulsating RSG model, whereas hydrostatic stellar models are not well-suited. The data coverage at early times is insufficient to break all degeneracies, but we find constraints on the explosion phase. We find no evidence for the claimed pulsation period of the SN~2024ggi progenitor, as it matches Spitzer's orbital period. This study underscores the importance of hydrodynamical pre-SN stellar models. It implies an important shift in our understanding of the last stages of massive star evolution, the interpretation of pre-SN properties, the connection between SNe and their progenitors, and the missing RSG problem.
|
[
{
"version": "v1",
"created": "Thu, 14 Aug 2025 21:57:10 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 15:43:21 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Laplace",
"Eva",
""
],
[
"Bronner",
"Vincent A.",
""
],
[
"Schneider",
"Fabian R. N.",
""
],
[
"Podsiadlowski",
"Philipp",
""
]
] |
2508.11133
|
Tomer Wolfson
|
Tomer Wolfson, Harsh Trivedi, Mor Geva, Yoav Goldberg, Dan Roth, Tushar Khot, Ashish Sabharwal, Reut Tsarfaty
|
MoNaCo: More Natural and Complex Questions for Reasoning Across Dozens of Documents
|
Accepted for publication in Transactions of the Association for Computational Linguistics (TACL), 2025. Authors pre-print
| null | null | null |
cs.CL cs.AI cs.DB
|
http://creativecommons.org/licenses/by/4.0/
|
Automated agents, powered by Large language models (LLMs), are emerging as the go-to tool for querying information. However, evaluation benchmarks for LLM agents rarely feature natural questions that are both information-seeking and genuinely time-consuming for humans. To address this gap we introduce MoNaCo, a benchmark of 1,315 natural and time-consuming questions that require dozens, and at times hundreds, of intermediate steps to solve -- far more than any existing QA benchmark. To build MoNaCo, we developed a decomposed annotation pipeline to elicit and manually answer real-world time-consuming questions at scale. Frontier LLMs evaluated on MoNaCo achieve at most 61.2% F1, hampered by low recall and hallucinations. Our results underscore the limitations of LLM-powered agents in handling the complexity and sheer breadth of real-world information-seeking tasks -- with MoNaCo providing an effective resource for tracking such progress. The MoNaCo benchmark, codebase, prompts and models predictions are all publicly available at: https://tomerwolgithub.github.io/monaco
|
[
{
"version": "v1",
"created": "Fri, 15 Aug 2025 00:58:10 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 17:03:15 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Wolfson",
"Tomer",
""
],
[
"Trivedi",
"Harsh",
""
],
[
"Geva",
"Mor",
""
],
[
"Goldberg",
"Yoav",
""
],
[
"Roth",
"Dan",
""
],
[
"Khot",
"Tushar",
""
],
[
"Sabharwal",
"Ashish",
""
],
[
"Tsarfaty",
"Reut",
""
]
] |
2508.11324
|
V. G. Gurzadyan
|
Vahe Gurzadyan, Arsen Bobokhyan
|
Vishap stelae as cult dedicated prehistoric monuments of Armenian Highlands: data analysis and interpretation
|
22 pages, 1 Table, 10 figs; Heritage Science (accepted for publication)
|
npj Heritage Science, 13, 435 (2025)
|
10.1038/s40494-025-01998-z
| null |
physics.soc-ph physics.hist-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vishaps, or dragon stones, are prehistoric stelae discovered in the high-altitude mountainous regions of modern-day Armenia and adjacent regions. The first statistical analysis of their elevation distribution and size reveals that their construction was intentionally labor-intensive rather than arbitrary. The findings support the hypothesis that vishaps were closely associated with an ancient water cult, as they are predominantly situated near water sources, including high-altitude springs and discovered prehistoric irrigation systems. Furthermore, the unexpected bimodal distribution of their altitudes suggests specific placement patterns, potentially linked to seasonal human activities or ritual practices. These findings contribute to a deeper understanding of the symbolic and functional significance of vishap stelae within the framework of prehistoric social and religious systems.
|
[
{
"version": "v1",
"created": "Fri, 15 Aug 2025 08:50:38 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Gurzadyan",
"Vahe",
""
],
[
"Bobokhyan",
"Arsen",
""
]
] |
2508.11622
|
Ziyuan Chen
|
Ziyuan Chen, Yifan Jiang, Jingyuan Liu, Fang Yao
|
Deconfounding via Profiled Transfer Learning
|
We need to further refine this paper
| null | null | null |
stat.ME
|
http://creativecommons.org/licenses/by/4.0/
|
Unmeasured confounders are a major source of bias in regression-based effect estimation and causal inference. In this paper, we advocate a new profiled transfer learning framework, ProTrans, to address confounding effects in the target dataset, when additional source datasets that possess similar confounding structures are available. We introduce the concept of profiled residuals to characterize the shared confounding patterns between source and target datasets. By incorporating these profiled residuals into the target debiasing step, we effectively mitigates the latent confounding effects. We also propose a source selection strategy to enhance robustness of ProTrans against noninformative sources. As a byproduct, ProTrans can also be utilized to estimate treatment effects when potential confounders exist, without the use of auxiliary features such as instrumental or proxy variables, which are often challenging to select in practice. Theoretically, we prove that the resulting estimated model shift from sources to target is confounding-free without any assumptions imposed on the true confounding structure, and that the target parameter estimation achieves the minimax optimal rate under mild conditions. Simulated and real-world experiments validate the effectiveness of ProTrans and support the theoretical findings.
|
[
{
"version": "v1",
"created": "Fri, 15 Aug 2025 17:48:36 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 10:52:21 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Chen",
"Ziyuan",
""
],
[
"Jiang",
"Yifan",
""
],
[
"Liu",
"Jingyuan",
""
],
[
"Yao",
"Fang",
""
]
] |
2508.12097
|
Christophe Galland
|
Zhiyuan Xie, Nobuaki Oyamada, Francesco Ciccarello, Wen Chen, Christophe Galland
|
Continuous-wave, high-resolution, ultra-broadband mid-infrared nonlinear spectroscopy with tunable plasmonic nanocavities
| null | null | null | null |
physics.optics cond-mat.mes-hall
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Vibrational sum- and difference-frequency generation (SFG and DFG) spectroscopy is a technique to probe the nonlinear response of interfaces at mid-infrared wavelengths while detecting the upconverted signal at visible wavelengths. It has recently transitioned from large-area molecular films and colloidal suspensions to nanoscale structures, enabled by dual-resonant plasmonic nanocavities that confine light and matter in the same deep-subwavelength sensing volume. Here, we implement high-resolution ($<1$~cm$^{-1}$), continuous-wave ultra-broadband SFG and DFG spectroscopy from $860$ to $1670~\mathrm{cm}^{-1}$ on a variety of dual-resonant antennas under ambient conditions. The use of a commercial, broadly tunable quantum-cascade laser and the absence of geometric phase-matching requirements speed up and simplify data acquisition. At the same time, simultaneous access to SFG and DFG sidebands permits ratiometric signal analysis, enabling both common-noise rejection and field-normalized characterization of coherent response. The resulting spectra harbour the signatures of coherent interference between the resonant (vibrational) and nonresonant (electronic) contributions to the $\chi^{(2)}$ susceptibility, previously only observed under fs- and ps-pulsed excitation. We demonstrate the versatility and reproducibility of our approach across several molecules that yield distinct resonant and nonresonant mid-infrared responses. Moreover, we identify robust differences between our data and DFT calculations that hint at the modification of the infrared molecular response by the nanocavity environment. Together, these features position our platform as a scalable route toward multiplexed, high-resolution mid-IR sensing and a potential basis for coherent single- or few-molecule vibrational spectroscopy at the chip level.
|
[
{
"version": "v1",
"created": "Sat, 16 Aug 2025 16:38:07 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 13:16:20 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Xie",
"Zhiyuan",
""
],
[
"Oyamada",
"Nobuaki",
""
],
[
"Ciccarello",
"Francesco",
""
],
[
"Chen",
"Wen",
""
],
[
"Galland",
"Christophe",
""
]
] |
2508.12232
|
Abbas Heydarnoori
|
Arshia Akhavan, Alireza Hosseinpour, Abbas Heydarnoori, Mehdi Keshani
|
LinkAnchor: An Autonomous LLM-Based Agent for Issue-to-Commit Link Recovery
| null | null | null | null |
cs.SE cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Issue-to-commit link recovery plays an important role in software traceability and improves project management. However, it remains a challenging task. A study on GitHub shows that only 42.2% of the issues are correctly linked to their commits. This highlights the potential for further development and research in this area. Existing studies have employed various AI/ML-based approaches, and with the recent development of large language models, researchers have leveraged LLMs to tackle this problem. These approaches suffer from two main issues. First, LLMs are constrained by limited context windows and cannot ingest all of the available data sources, such as long commit histories, extensive issue comments, and large code repositories. Second, most methods operate on individual issue-commit pairs; that is, given a single issue-commit pair, they determine whether the commit resolves the issue. This quickly becomes impractical in real-world repositories containing tens of thousands of commits. To address these limitations, we present LinkAnchor, the first autonomous LLM-based agent designed for issue-to-commit link recovery. The lazy-access architecture of LinkAnchor enables the underlying LLM to access the rich context of software, spanning commits, issue comments, and code files, without exceeding the token limit by dynamically retrieving only the most relevant contextual data. Additionally, LinkAnchor is able to automatically pinpoint the target commit rather than exhaustively scoring every possible candidate. Our evaluations show that LinkAnchor outperforms state-of-the-art issue-to-commit link recovery approaches by 60-262% in Hit@1 score across all our case study projects. We also publicly release LinkAnchor as a ready-to-use tool, along with our replication package. LinkAnchor is designed and tested for GitHub and Jira, and is easily extendable to other platforms.
|
[
{
"version": "v1",
"created": "Sun, 17 Aug 2025 04:21:44 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 23:35:13 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Akhavan",
"Arshia",
""
],
[
"Hosseinpour",
"Alireza",
""
],
[
"Heydarnoori",
"Abbas",
""
],
[
"Keshani",
"Mehdi",
""
]
] |
2508.12600
|
Manuel Barrientos
|
Manuel Barrientos, Mukremin Kilic, Simon Blouin, Michael R. Hayden, Sanjib Sharma, and Matthew J. Green
|
The Fraction of Distilled White Dwarfs with Long-Lived Habitable Zones
|
Accepted for Publication in ApJL. 7 pages, 4 figures
|
The Astrophysical Journal Letters, Volume 990, Year 2025, Issue 2, L47
|
10.3847/2041-8213/adfc4a
| null |
astro-ph.SR astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
After carbon and oxygen, $^{22}$Ne is the most abundant element in white dwarf interiors. As C/O white dwarfs (WDs) crystallize, they are predicted to go through a distillation process in the central layers if they have a sufficiently high $^{22}$Ne mass fraction of $\gtrsim2.5$\%. Observational evidence for distillation comes from an over-density of WDs on the Q-branch in Gaia color-magnitude diagrams, which indicates that $\sim6$\% of massive WDs are delayed in their cooling by as much as $\sim10$ Gyr. However, it is unclear how these stars end up with such a high concentration of $^{22}$Ne and if a significant fraction of the more common average-mass WDs go through distillation. We argue that a significant metal-rich stellar population in the solar neighborhood should lead to distilled WDs, without requiring a binary merger. We use MESA along with the CNO abundances derived from high-resolution spectroscopy of stars included in the Hypatia catalog to predict the $^{22}$Ne mass fraction in their descendant WDs. We find that 0.6-2.5\% of the WDs in the solar neighborhood have sufficient $^{22}$Ne in their interiors to go through multi-Gyr cooling delays, which could significantly inflate their numbers in the observed samples. Hence, $^{22}$Ne distillation and long-lived habitable zones around WDs should be relatively common in the solar neighborhood. We also use a Galactic model to predict the fraction of WDs that go through distillation as a function of Galactocentric distance. The fraction of distilled WDs is $\sim2$-8\% near the Galactic center, and declines steadily toward the outer disk.
|
[
{
"version": "v1",
"created": "Mon, 18 Aug 2025 03:26:32 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Barrientos",
"Manuel",
""
],
[
"Kilic",
"Mukremin",
""
],
[
"Blouin",
"Simon",
""
],
[
"Hayden",
"Michael R.",
""
],
[
"Sharma",
"Sanjib",
""
],
[
"Green",
"Matthew J.",
""
]
] |
2508.12750
|
Hao Cheng
|
Linhao Li, Boya Jin, Zizhe Li, Lanqing Guo, Hao Cheng, Bo Li, Yongfeng Dong
|
D2-Mamba: Dual-Scale Fusion and Dual-Path Scanning with SSMs for Shadow Removal
|
Paper Under Review
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Shadow removal aims to restore images that are partially degraded by shadows, where the degradation is spatially localized and non-uniform. Unlike general restoration tasks that assume global degradation, shadow removal can leverage abundant information from non-shadow regions for guidance. However, the transformation required to correct shadowed areas often differs significantly from that of well-lit regions, making it challenging to apply uniform correction strategies. This necessitates the effective integration of non-local contextual cues and adaptive modeling of region-specific transformations. To this end, we propose a novel Mamba-based network featuring dual-scale fusion and dual-path scanning to selectively propagate contextual information based on transformation similarity across regions. Specifically, the proposed Dual-Scale Fusion Mamba Block (DFMB) enhances multi-scale feature representation by fusing original features with low-resolution features, effectively reducing boundary artifacts. The Dual-Path Mamba Group (DPMG) captures global features via horizontal scanning and incorporates a mask-aware adaptive scanning strategy, which improves structural continuity and fine-grained region modeling. Experimental results demonstrate that our method significantly outperforms existing state-of-the-art approaches on shadow removal benchmarks.
|
[
{
"version": "v1",
"created": "Mon, 18 Aug 2025 09:20:21 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:41:31 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Li",
"Linhao",
""
],
[
"Jin",
"Boya",
""
],
[
"Li",
"Zizhe",
""
],
[
"Guo",
"Lanqing",
""
],
[
"Cheng",
"Hao",
""
],
[
"Li",
"Bo",
""
],
[
"Dong",
"Yongfeng",
""
]
] |
2508.13073
|
Wei Li
|
Rui Shao, Wei Li, Lingsen Zhang, Renshan Zhang, Zhiyang Liu, Ran Chen, Liqiang Nie
|
Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Survey
|
Project Page: https://github.com/JiuTian-VL/Large-VLM-based-VLA-for-Robotic-Manipulation
| null | null | null |
cs.RO cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Robotic manipulation, a key frontier in robotics and embodied AI, requires precise motor control and multimodal understanding, yet traditional rule-based methods fail to scale or generalize in unstructured, novel environments. In recent years, Vision-Language-Action (VLA) models, built upon Large Vision-Language Models (VLMs) pretrained on vast image-text datasets, have emerged as a transformative paradigm. This survey provides the first systematic, taxonomy-oriented review of large VLM-based VLA models for robotic manipulation. We begin by clearly defining large VLM-based VLA models and delineating two principal architectural paradigms: (1) monolithic models, encompassing single-system and dual-system designs with differing levels of integration; and (2) hierarchical models, which explicitly decouple planning from execution via interpretable intermediate representations. Building on this foundation, we present an in-depth examination of large VLM-based VLA models: (1) integration with advanced domains, including reinforcement learning, training-free optimization, learning from human videos, and world model integration; (2) synthesis of distinctive characteristics, consolidating architectural traits, operational strengths, and the datasets and benchmarks that support their development; (3) identification of promising directions, including memory mechanisms, 4D perception, efficient adaptation, multi-agent cooperation, and other emerging capabilities. This survey consolidates recent advances to resolve inconsistencies in existing taxonomies, mitigate research fragmentation, and fill a critical gap through the systematic integration of studies at the intersection of large VLMs and robotic manipulation. We provide a regularly updated project page to document ongoing progress: https://github.com/JiuTian-VL/Large-VLM-based-VLA-for-Robotic-Manipulation
|
[
{
"version": "v1",
"created": "Mon, 18 Aug 2025 16:45:48 GMT"
},
{
"version": "v2",
"created": "Mon, 01 Sep 2025 08:10:01 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Shao",
"Rui",
""
],
[
"Li",
"Wei",
""
],
[
"Zhang",
"Lingsen",
""
],
[
"Zhang",
"Renshan",
""
],
[
"Liu",
"Zhiyang",
""
],
[
"Chen",
"Ran",
""
],
[
"Nie",
"Liqiang",
""
]
] |
2508.13401
|
Andrei Dumitriu
|
Andrei Dumitriu, Florin Miron, Florin Tatui, Radu Tudor Ionescu, Radu Timofte, Aakash Ralhan, Florin-Alexandru Vasluianu, Shenyang Qian, Mitchell Harley, Imran Razzak, Yang Song, Pu Luo, Yumei Li, Cong Xu, Jinming Chai, Kexin Zhang, Licheng Jiao, Lingling Li, Siqi Yu, Chao Zhang, Kehuan Song, Fang Liu, Puhua Chen, Xu Liu, Jin Hu, Jinyang Xu, Biao Liu
|
AIM 2025 Rip Current Segmentation (RipSeg) Challenge Report
|
Challenge report paper from AIM2025 Workshop at ICCVW 2025
| null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This report presents an overview of the AIM 2025 RipSeg Challenge, a competition designed to advance techniques for automatic rip current segmentation in still images. Rip currents are dangerous, fast-moving flows that pose a major risk to beach safety worldwide, making accurate visual detection an important and underexplored research task. The challenge builds on RipVIS, the largest available rip current dataset, and focuses on single-class instance segmentation, where precise delineation is critical to fully capture the extent of rip currents. The dataset spans diverse locations, rip current types, and camera orientations, providing a realistic and challenging benchmark.
In total, $75$ participants registered for this first edition, resulting in $5$ valid test submissions. Teams were evaluated on a composite score combining $F_1$, $F_2$, $AP_{50}$, and $AP_{[50:95]}$, ensuring robust and application-relevant rankings. The top-performing methods leveraged deep learning architectures, domain adaptation techniques, pretrained models, and domain generalization strategies to improve performance under diverse conditions.
This report outlines the dataset details, competition framework, evaluation metrics, and final results, providing insights into the current state of rip current segmentation. We conclude with a discussion of key challenges, lessons learned from the submissions, and future directions for expanding RipSeg.
|
[
{
"version": "v1",
"created": "Mon, 18 Aug 2025 23:34:56 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 17:20:31 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Dumitriu",
"Andrei",
""
],
[
"Miron",
"Florin",
""
],
[
"Tatui",
"Florin",
""
],
[
"Ionescu",
"Radu Tudor",
""
],
[
"Timofte",
"Radu",
""
],
[
"Ralhan",
"Aakash",
""
],
[
"Vasluianu",
"Florin-Alexandru",
""
],
[
"Qian",
"Shenyang",
""
],
[
"Harley",
"Mitchell",
""
],
[
"Razzak",
"Imran",
""
],
[
"Song",
"Yang",
""
],
[
"Luo",
"Pu",
""
],
[
"Li",
"Yumei",
""
],
[
"Xu",
"Cong",
""
],
[
"Chai",
"Jinming",
""
],
[
"Zhang",
"Kexin",
""
],
[
"Jiao",
"Licheng",
""
],
[
"Li",
"Lingling",
""
],
[
"Yu",
"Siqi",
""
],
[
"Zhang",
"Chao",
""
],
[
"Song",
"Kehuan",
""
],
[
"Liu",
"Fang",
""
],
[
"Chen",
"Puhua",
""
],
[
"Liu",
"Xu",
""
],
[
"Hu",
"Jin",
""
],
[
"Xu",
"Jinyang",
""
],
[
"Liu",
"Biao",
""
]
] |
2508.13613
|
Hoseob Seo
|
Hoseob Seo
|
On contact Hamiltonian functions of singular contact structures
| null | null | null | null |
math.DG math.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
For contact manifolds, it is well-known that the map which assigns to an infinitesimal contact transformation its contact Hamiltonian function is a linear isomorphism, and an explicit local formula for its inverse can be given. In contrast, infinitesimal contact transformations on a singular contact structure have not been well-studied yet. In this article, we show that this map is injective and present an explicit local formula for its inverse when the contact structure has singularities of the first type with structurally smooth Martinet hypersurface.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 08:17:53 GMT"
},
{
"version": "v2",
"created": "Sat, 30 Aug 2025 01:42:46 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Seo",
"Hoseob",
""
]
] |
2508.13705
|
Junlei Mu
|
Junlei Mu, Hong Zhang, Xing Ji, Yang Zhang, Gang Chen, Kun Xu
|
Very High-order Compact Gas-kinetic Scheme With Discontinuity Feedback Factor
| null | null | null | null |
physics.comp-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents a robust and efficient very high-order scheme for compressible flow simulation, addressing critical limitations of existing high-order methods. The proposed scheme combines the compact gas-kinetic scheme (CGKS) with an adaptive stencil extension reconstruction with discontinuity feedback factor (ASE-DFF), achieving significant improvements in both robustness and computational efficiency. Traditional weighted essentially non-oscillatory (WENO) schemes suffer from reduced robustness at higher order and require costly smoothness indicators for large stencils. Meanwhile, compact methods based on Discontinuous Galerkin (DG) and Flux Reconstruction (FR) struggle with poor time-marching efficiency. In contrast, the ASE-DFF-CGKS introduces two key innovations: (1) a unified framework enabling arbitrarily high-order compact gas-kinetic scheme without sacrificing large CFL number, and (2) a discontinuity feedback factor that eliminates the need for expensive smoothness indicator calculations while essentially keeping first-order robustness near discontinuities. The scheme's advantages are demonstrated through benchmark simulations. It maintains a CFL number above 0.5 for up to 9th-order case, unlike conventional compact methods that restrict a CFL less than 0.05. Also it delivers high-resolution results for flow involving strong shock and rarefaction wave. This work provides a practically impactful solution for high-fidelity compressible flow simulation, balancing computational efficiency, high-order accuracy and robustness in challenging flow regimes.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 10:13:25 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:50:52 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Mu",
"Junlei",
""
],
[
"Zhang",
"Hong",
""
],
[
"Ji",
"Xing",
""
],
[
"Zhang",
"Yang",
""
],
[
"Chen",
"Gang",
""
],
[
"Xu",
"Kun",
""
]
] |
2508.14179
|
Jorge L. Rosa-Ra\'ices
|
Jorge L. Rosa-Ra\'ices and David T. Limmer
|
Nonadiabatic force matching for alchemical free-energy estimation
|
Implemented nomenclature changes to refer to "flow-matching models" instead of "diffusion-denoising models" throughout the article
| null | null | null |
physics.chem-ph cond-mat.stat-mech physics.bio-ph
|
http://creativecommons.org/licenses/by/4.0/
|
We propose a method to compute free-energy differences from nonadiabatic alchemical transformations using flow-based generative models. The method, nonadiabatic force matching, hinges on estimating the dissipation along an alchemical switching process in terms of a nonadiabatic force field that can be learned through stochastic flow matching. The learned field can be used in conjunction with short-time trajectory data to evaluate upper and lower bounds on the alchemical free energy that variationally converge to the exact value if the field is optimal. Applying the method to evaluate the alchemical free energy of atomistic models shows that it can substantially reduce the simulation cost of a free-energy estimate at negligible loss of accuracy when compared with thermodynamic integration.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 18:07:43 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 20:28:06 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Rosa-Raíces",
"Jorge L.",
""
],
[
"Limmer",
"David T.",
""
]
] |
2508.14376
|
Xuzhou Zhan
|
Xuzhou Zhan and Zixiang Ni
|
A generalized Hurwitz stability criterion via rectangular block Hankel matrices for nonmonic matrix polynomials
| null | null | null | null |
math.OC cs.NA math.NA
|
http://creativecommons.org/licenses/by/4.0/
|
We develop a Hurwitz stability criterion for nonmonic matrix polynomials via column reduction, generalizing existing approaches constrained by the monic assumption, as well as Gantmacher's classical stability criterion via Markov parameters. Starting from redefining the associated Markov parameters through a column-wise adaptive splitting method, our framework constructs two structured matrices whose rectangular Hankel blocks are obtained via the extraction of these parameters. We establish an explicit interrelation between the inertias of column reduced matrix polynomials and the derived structured matrices. Furthermore, we demonstrate that the Hurwitz stability of column reduced matrix polynomials can be determined by the Hermitian positive definiteness of these rectangular block Hankel matrices.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 03:06:10 GMT"
},
{
"version": "v2",
"created": "Mon, 25 Aug 2025 01:50:45 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 14:14:16 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Zhan",
"Xuzhou",
""
],
[
"Ni",
"Zixiang",
""
]
] |
2508.14546
|
Yasunori Lee
|
Yuya O. Nakagawa, Yasunori Lee
|
Application of resource theory based on free Clifford+kT computation to early fault-tolerant quantum computing
|
27+1 pages; v2: references added
| null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent advances in quantum hardware are bringing fault-tolerant quantum computing (FTQC) closer to reality. In the early stage of FTQC, however, the numbers of available logical qubits and high-fidelity $T$ gates remain limited, making it crucial to optimize the quantum resource usage. In this work, we aim to study the simulation cost of general quantum states under the constraint that only $k$ $T$ gates can be used, alongside an unlimited number of Clifford gates. Inspired by the notion of robustness of magic (RoM) which quantifies the cost of quantum-circuit simulation using stabilizer states ($k = 0$), we introduce its generalization, which we call Clifford+$kT$ robustness, treating Clifford+$kT$ states as free resources. We explore theoretical properties of Clifford+$kT$ robustness and in particular derive a lower bound that reveals the (in)efficiency of quantum-circuit simulation using Clifford+$kT$ states. Through numerical computations, we also evaluate Clifford+$kT$ robustness for key resource states for universal quantum computation, such as tensor products of the magic states. Our results allow to assess the sampling-cost reduction achieved by the use of Clifford+$kT$ states instead of stabilizer states, providing practical guidance for efficient resource usage in the early-FTQC era.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 09:00:00 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 09:00:00 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Nakagawa",
"Yuya O.",
""
],
[
"Lee",
"Yasunori",
""
]
] |
2508.14966
|
Matthew Forslund
|
Luigi Bellafronte, Sally Dawson, Clara Del Pio, Matthew Forslund, Pier Paolo Giardino
|
Complete NLO SMEFT Electroweak Corrections to Higgs Decays
|
6 pages, 4 figures; v2: Treatment of RGE logs improved
| null | null |
IFT-UAM/CSIC-25-86
|
hep-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Precise predictions for Higgs decays are a crucial ingredient of the search for beyond the Standard Model (BSM) physics and the Standard Model Effective Field Theory (SMEFT) is a valuable tool for quantifying deviations from the Standard Model (SM). We present the complete set of predictions for the 2- and 3- body Higgs decays at next-to-leading order (NLO), considering QCD and electroweak corrections and including all contributions from the dimension-6 SMEFT operators and with an arbitrary flavor structure. Including the NLO SMEFT results for Higgs decays greatly increases the sensitivity to BSM physics of the $e^+e^-\rightarrow Zh$ process at FCC-ee, as compared with that obtained using only the total cross section.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 18:00:01 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 23:15:50 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Bellafronte",
"Luigi",
""
],
[
"Dawson",
"Sally",
""
],
[
"Del Pio",
"Clara",
""
],
[
"Forslund",
"Matthew",
""
],
[
"Giardino",
"Pier Paolo",
""
]
] |
2508.15060
|
Wen-Tai Hsu
|
Wen-Tai Hsu
|
Asymptotic analysis on narrow tubes: narrow escape problems and diffusion processes
| null | null | null | null |
math.PR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper investigates a diffusion process in a narrow tubular domain with reflecting boundary conditions, where the geometry serves as a singular perturbation of an underlying graph in $\mathbb{R}^2$ or $\mathbb{R}^3$. The construction incorporates distinct scaling regimes in the neighborhoods of the graph's vertices and edges. We show that, in the limit, the projected process converges weakly to a diffusion process on the graph, with gluing conditions at the vertices that depend on the relative scales of the neighborhoods. Our analysis relies on a detailed understanding of the narrow escape problem in domains with bottlenecks. In particular, we rigorously derive the asymptotic behavior of the expected escape time, establish the asymptotic exponential distribution of escape times and obtain exit place estimates, results that may be of independent interest.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 20:48:17 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 05:35:01 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Hsu",
"Wen-Tai",
""
]
] |
2508.15075
|
Simone Hagey
|
Simone R. Hagey, Billy Edwards, Angelos Tsiaras, Aaron C. Boley, Anastasia Kokori, Norio Narita, Pedro V. Sada, Filip Walter, Robert T. Zellem, Napaporn A-thano, Kevin B. Alton, Miguel \'Angel \'Alava Amat, Paul Benni, Emmanuel Besson, Patrick Brandebourg, Marc Bretton, Mauro Cal\'o, Martin Valentine Crow, Jean-Christophe Dalouzy, Marc Deldem, T\~onis Eenm\"ae, Stephane Ferratfiat, Pere Guerra, Gary Vander Haagen, Ken Hose, Adrian Jones, Yves Jongen, Didier Laloum, Stefano Lora, Alessandro Marchini, Jacques Michelet, Matej Mihel\v{c}i\v{c}, Johannes Mieglitz, Eric Miny, David Molina, Mario Morales Aimar, Raphael Nicollerat, Ivo Peretto, Manfred Raetz, Fran\c{c}ois Regembal, Robert Roth, Lionel Rousselot, Mark Salisbury, Darryl Sergison, Ana\"el W\"unsche, Jaroslav Trnka
|
TrES-1 b: A Case Study in Detecting Secular Evolution of Exoplanet Orbits
|
32 pages, 12 figures, 10 tables, accepted to AJ
| null |
10.3847/1538-3881/aded15
| null |
astro-ph.EP
|
http://creativecommons.org/licenses/by/4.0/
|
We present a comprehensive analysis of transit, eclipse, and radial velocity data of the hot Jupiter TrES-1 b and confirm evidence of orbital variations on secular timescales. Apparent variations due to systemic motion and light travel time effects have been ruled out, indicating that the observed changes are dynamical in origin. Joint modeling of the TrES-1 b data favors an apsidal precession model, but the rapid precession rate of $4^\circ$ yr$^{-1}$ cannot be explained without invoking an undetected close-in planetary companion, which remains unseen in the data. While radial velocity measurements reveal a previously undetected companion candidate on a wide, eccentric orbit, it is unlikely to drive the observed evolution of TrES-1 b. However, an orbital decay model provides a plausible alternative if the loss of orbital energy is driven by planetary obliquity tides. We find that the best-fit orbital decay rate of $-7.1^{ +1.5}_{-1.6}$ ms yr$^{-1}$ is aligned with theoretical predictions for modified tidal quality factors of hot Jupiters if TrES-1 b has a planetary obliquity $\varepsilon_p > 30^\circ$. We encourage follow-up observations of this system, particularly of eclipse timing and radial velocities, to further constrain the nature of the observed evolution. This paper provides a practical framework for studying secular variations and aims to accelerate future research on similar systems.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 21:18:51 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Hagey",
"Simone R.",
""
],
[
"Edwards",
"Billy",
""
],
[
"Tsiaras",
"Angelos",
""
],
[
"Boley",
"Aaron C.",
""
],
[
"Kokori",
"Anastasia",
""
],
[
"Narita",
"Norio",
""
],
[
"Sada",
"Pedro V.",
""
],
[
"Walter",
"Filip",
""
],
[
"Zellem",
"Robert T.",
""
],
[
"A-thano",
"Napaporn",
""
],
[
"Alton",
"Kevin B.",
""
],
[
"Amat",
"Miguel Ángel Álava",
""
],
[
"Benni",
"Paul",
""
],
[
"Besson",
"Emmanuel",
""
],
[
"Brandebourg",
"Patrick",
""
],
[
"Bretton",
"Marc",
""
],
[
"Caló",
"Mauro",
""
],
[
"Crow",
"Martin Valentine",
""
],
[
"Dalouzy",
"Jean-Christophe",
""
],
[
"Deldem",
"Marc",
""
],
[
"Eenmäe",
"Tõnis",
""
],
[
"Ferratfiat",
"Stephane",
""
],
[
"Guerra",
"Pere",
""
],
[
"Haagen",
"Gary Vander",
""
],
[
"Hose",
"Ken",
""
],
[
"Jones",
"Adrian",
""
],
[
"Jongen",
"Yves",
""
],
[
"Laloum",
"Didier",
""
],
[
"Lora",
"Stefano",
""
],
[
"Marchini",
"Alessandro",
""
],
[
"Michelet",
"Jacques",
""
],
[
"Mihelčič",
"Matej",
""
],
[
"Mieglitz",
"Johannes",
""
],
[
"Miny",
"Eric",
""
],
[
"Molina",
"David",
""
],
[
"Aimar",
"Mario Morales",
""
],
[
"Nicollerat",
"Raphael",
""
],
[
"Peretto",
"Ivo",
""
],
[
"Raetz",
"Manfred",
""
],
[
"Regembal",
"François",
""
],
[
"Roth",
"Robert",
""
],
[
"Rousselot",
"Lionel",
""
],
[
"Salisbury",
"Mark",
""
],
[
"Sergison",
"Darryl",
""
],
[
"Wünsche",
"Anaël",
""
],
[
"Trnka",
"Jaroslav",
""
]
] |
2508.15318
|
Qian-Rui Lee
|
Qian-Rui Lee, Daw-Wei Wang
|
Flow Matching at Scale: A Machine Learning Framework for Efficient Large-Size Sampling of Many-Body Systems
| null | null | null | null |
cond-mat.stat-mech cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We propose a machine learning framework based on Flow Matching to overcome the scaling limitations of Markov Chain Monte Carlo (MCMC) methods. We demonstrate its capability in the 2D XY model, where a single network, trained only on configurations from a small ($32\times 32$) lattice at sparse temperature points, generates reliable samples for a significantly larger system ($128\times 128$) across a continuous temperature range without retraining. The generated configurations show strong agreement with key thermodynamic observables and correctly capture the signatures of the Berezinskii-Kosterlitz-Thouless (BKT) transition. This dual generalization is enabled by the Flow Matching framework, which allows us to learn a continuous, temperature-conditioned mapping. At the same time, the inductive biases of the underlying CNN architecture ensure that the learned local physical rules are scale-invariant. This "train-small, generate-large" capability offers a powerful and efficient alternative for studying critical phenomena. The method can be directly applied to other classical or quantum many-body systems described by continuous fields on a lattice. Furthermore, this framework can serve as a powerful proposal generator in a hybrid scheme with MCMC, dramatically accelerating high-precision studies of the thermodynamic limit.
|
[
{
"version": "v1",
"created": "Thu, 21 Aug 2025 07:29:25 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 17:54:29 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lee",
"Qian-Rui",
""
],
[
"Wang",
"Daw-Wei",
""
]
] |
2508.15327
|
Xiancheng Gao
|
Xiancheng Gao, Yufeng Shi, Wengang Zhou and Houqiang Li
|
Search-Based Credit Assignment for Offline Preference-Based Reinforcement Learning
|
7 pages, 6 figures, under review
| null | null | null |
cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Offline reinforcement learning refers to the process of learning policies from fixed datasets, without requiring additional environment interaction. However, it often relies on well-defined reward functions, which are difficult and expensive to design. Human feedback is an appealing alternative, but its two common forms, expert demonstrations and preferences, have complementary limitations. Demonstrations provide stepwise supervision, but they are costly to collect and often reflect limited expert behavior modes. In contrast, preferences are easier to collect, but it is unclear which parts of a behavior contribute most to a trajectory segment, leaving credit assignment unresolved. In this paper, we introduce a Search-Based Preference Weighting (SPW) scheme to unify these two feedback sources. For each transition in a preference labeled trajectory, SPW searches for the most similar state-action pairs from expert demonstrations and directly derives stepwise importance weights based on their similarity scores. These weights are then used to guide standard preference learning, enabling more accurate credit assignment that traditional approaches struggle to achieve. We demonstrate that SPW enables effective joint learning from preferences and demonstrations, outperforming prior methods that leverage both feedback types on challenging robot manipulation tasks.
|
[
{
"version": "v1",
"created": "Thu, 21 Aug 2025 07:41:45 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:20:31 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Gao",
"Xiancheng",
""
],
[
"Shi",
"Yufeng",
""
],
[
"Zhou",
"Wengang",
""
],
[
"Li",
"Houqiang",
""
]
] |
2508.16664
|
Sana Khalil
|
Sana Khalil and Angela Warner
|
Invisible Labor, Visible Barriers: The Socioeconomic Realities of Women's Work in Pakistan
| null | null | null | null |
econ.GN q-fin.EC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We highlight the barriers shaping women's economic opportunities in Pakistan, where female labor force participation remains among the lowest globally. Labor force surveys (2020-21) show a stark rural-urban divide: 28 percent for rural women versus 69 percent for rural men, and 10 percent for urban women versus 66 percent for urban men. Unemployment is higher for women (7 percent in rural areas; 16 percent in urban areas) than for men (5 and 6 percent, respectively). Women are concentrated in agriculture (68 percent), with limited presence in services (17 percent) and industry (15 percent), and mostly in rural (51 percent) or home-based (30 percent) work; only 14 percent are in formal business settings. Employment status reflects vulnerability: 63 percent of rural women are unpaid contributing family workers versus 17 percent of urban women. Interviews with married women in Karachi underscore childcare constraints, harassment and safety concerns, transport barriers, and family opposition. Together, the evidence points to structural and cultural constraints that restrict access to paid work; easing them will require labor market reforms, better transport and childcare, stronger protections against harassment and discrimination, and a gradual change in gender norms and household decision-making.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 21:08:48 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 19:46:40 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Khalil",
"Sana",
""
],
[
"Warner",
"Angela",
""
]
] |
2508.17327
|
Wenjian Lu
|
Run Lv, Wenqian Tu, Dingfu Shao, Yuping Sun, and Wenjian Lu
|
Orbital-selective two-gap superconductivity in kagome metal CsV3Sb5
| null | null | null | null |
cond-mat.supr-con cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Recent experiments have revealed anisotropic multi-gap superconductivity in the kagome metal CsV3Sb5. However, the impact of multi-orbital character and electron-phonon coupling (EPC) on the multiple superconducting gaps remains not fully understood. In this work, we theoretically investigate the multi-orbital electronic structure and superconducting gap properties of CsV3Sb5 by combining first-principles calculations with superconducting density functional theory (SCDFT). Our results demonstrate that orbital-selective pairing drives the observed two-gap superconductivity in CsV3Sb5. Specifically, the two distinct gaps exhibit strong orbital dependence: a large, highly anisotropic gap (average magnitude ~0.64 meV) primarily originates from V-3d orbitals, while a small, isotropic gap (~0.25 meV) is associated with the Sb-5pz orbital. The V-3d orbitals exhibit strong coupling to the in-plane V-V bond-stretching and out-of-plane V-Sb bending phonon modes, whereas the Sb-5pz orbitals show weak coupling to the out-of-plane vibrations of both Cs and the apical Sb atoms. These findings provide fundamental insights into the orbital-selective superconductivity and EPC mechanisms in kagome CsV3Sb5.
|
[
{
"version": "v1",
"created": "Sun, 24 Aug 2025 12:20:50 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 01:53:57 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lv",
"Run",
""
],
[
"Tu",
"Wenqian",
""
],
[
"Shao",
"Dingfu",
""
],
[
"Sun",
"Yuping",
""
],
[
"Lu",
"Wenjian",
""
]
] |
2508.17518
|
Stefanos Chaliasos
|
Thomas Gassmann, Stefanos Chaliasos, Thodoris Sotiropoulos, Zhendong Su
|
Evaluating Compiler Optimization Impacts on zkVM Performance
| null | null | null | null |
cs.PF cs.PL
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Zero-knowledge proofs (ZKPs) are the cornerstone of programmable cryptography. They enable (1) privacy-preserving and verifiable computation across blockchains, and (2) an expanding range of off-chain applications such as credential schemes. Zero-knowledge virtual machines (zkVMs) lower the barrier by turning ZKPs into a drop-in backend for standard compilation pipelines. This lets developers write proof-generating programs in conventional languages (e.g., Rust or C++) instead of hand-crafting arithmetic circuits. However, these VMs inherit compiler infrastructures tuned for traditional architectures rather than for proof systems. In particular, standard compiler optimizations assume features that are absent in zkVMs, including cache locality, branch prediction, or instruction-level parallelism. Therefore, their impact on proof generation is questionable.
We present the first systematic study of the impact of compiler optimizations on zkVMs. We evaluate 64 LLVM passes, six standard optimization levels, and an unoptimized baseline across 58 benchmarks on two RISC-V-based zkVMs (RISC Zero and SP1). While standard LLVM optimization levels do improve zkVM performance (over 40\%), their impact is far smaller than on traditional CPUs, since their decisions rely on hardware features rather than proof constraints. Guided by a fine-grained pass-level analysis, we~\emph{slightly} refine a small set of LLVM passes to be zkVM-aware, improving zkVM execution time by up to 45\% (average +4.6\% on RISC Zero, +1\% on SP1) and achieving consistent proving-time gains. Our work highlights the potential of compiler-level optimizations for zkVM performance and opens new direction for zkVM-specific passes, backends, and superoptimizers.
|
[
{
"version": "v1",
"created": "Sun, 24 Aug 2025 20:51:06 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Gassmann",
"Thomas",
""
],
[
"Chaliasos",
"Stefanos",
""
],
[
"Sotiropoulos",
"Thodoris",
""
],
[
"Su",
"Zhendong",
""
]
] |
2508.17523
|
Royal Ihuaenyi PhD
|
Royal C. Ihuaenyi, Hongbo Zhao, Ruqing Fang, Ruobing Bai, Martin Z. Bazant, and Juner Zhu
|
Learning Reaction-Diffusion Kinetics from Mechanical Information
| null | null | null | null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
A central challenge in materials science is characterizing chemical processes that are elusive to direct measurement, particularly in functional materials operating under realistic conditions. Here, we demonstrate that mechanical strain fields contain sufficient information to reconstruct hidden chemical kinetics in coupled chemomechanical systems. Our partial differential equation-constrained learning framework decodes concentration-dependent diffusion kinetics, thermodynamic driving forces, and spatially heterogeneous reaction rates solely from mechanical observations. Using battery electrode materials as a model system, we demonstrate that the framework can accurately identify complex constitutive laws governing three distinct scenarios: classical Fickian diffusion, spinodal decomposition with pattern formation, and heterogeneous electrochemical reactions with spatial rate variations. The approach demonstrates robustness while maintaining accuracy with limited spatial data and reasonable experimental noise levels. Most significantly, the framework simultaneously infers multiple fundamental processes and properties, including diffusivity, reaction kinetics, chemical potential, and spatial heterogeneity maps, all from mechanical information alone. This method establishes a paradigm for materials characterization, enabling accurate learning of chemical processes in energy storage systems, catalysts, and phase-change materials where conventional diagnostics prove difficult. By revealing that mechanical deformation patterns serve as information-rich fingerprints of the underlying chemical processes, this work follows the pathway of inversely learning constitutive laws, with broad implications in materials science and engineering.
|
[
{
"version": "v1",
"created": "Sun, 24 Aug 2025 21:08:29 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 15:54:09 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Ihuaenyi",
"Royal C.",
""
],
[
"Zhao",
"Hongbo",
""
],
[
"Fang",
"Ruqing",
""
],
[
"Bai",
"Ruobing",
""
],
[
"Bazant",
"Martin Z.",
""
],
[
"Zhu",
"Juner",
""
]
] |
2508.17552
|
Ruy Exel
|
Ruy Exel
|
Consonant inverse semigroups
|
Minor changes
| null | null | null |
math.OA
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
We study necessary and sufficient conditions for two inverse semigroups to possess identical tight groupoids from the point of view of their algebraic, topological, and spectral order structures. The spectral order is a partial order relation that presents itself very naturally on the tight groupoid of an inverse semigroup and is related to the subtle difference between tight filters and ultra-filters. Up to a small glitch, the spectral order makes tight groupoids into ordered groupoids in Ehresmann's sense.
We introduce the notions of tight injectivity and tight surjectivity for inverse semigroup homomorphisms and show that, together, they provide necessary and sufficient conditions for the induced map to be an isomorphism of ordered topological groupoids. A homomorphism is then called a consonance provided these conditions are met.
A consonance is not necessarily injective or surjective, so it is likely not an invertible map. Due to this fact, the existence of a consonance between inverse semigroups does not define an equivalence relation, so we consider instead the equivalence relation it generates. When it applies to a pair of inverse semigroups we say that they are consonant.
The first main result of the paper shows that two inverse semigroups $S_1$ and $S_2$ are consonant if and only if their tight groupoids are isomorphic as ordered topological groupoids, if and only if there exists another inverse semigroup $T$ and consonances from each $S_i$ to $T$.
We also prove that, given an inverse semigroup $S$, there exists a largest inverse semigroup consonant to $S$, denoted $S^\tau$, and called the tight envelope of $S$. In a nutshell, $S^\tau$ is the inverse semigroup formed by the compact up-slices (i.e.~slices that are up-sets relative to the spectral order) in the tight groupoid of $S$. Among other thi ... (cut short by the system)
|
[
{
"version": "v1",
"created": "Sun, 24 Aug 2025 23:27:11 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 20:48:07 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Exel",
"Ruy",
""
]
] |
2508.17725
|
Hiroyuki Fujioka
|
Hiroyuki Fujioka and Takashi Higuchi
|
Impact of a Reflecting Material on a Search for Neutron--Antineutron Oscillations using Ultracold Neutrons
|
12 pages, 4 figures
| null | null | null |
hep-ex quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We investigate neutron--antineutron oscillations of ultracold neutrons in a storage bottle represented by a one-dimensional potential. The experimental sensitivity is determined by the annihilation rate of antineutrons. Its dependence on the antineutron reflectivity and the relative phase shift between the neutron and the antineutron wavefunctions by a reflection from the wall is derived. Optimization of the antineutron pseudopotential was found crucial to maximize the sensitivity of the experiment. Furthermore, methods are discussed for determining the antineutron pseudopotential, which has only been studied indirectly thus far.
|
[
{
"version": "v1",
"created": "Mon, 25 Aug 2025 07:06:32 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Fujioka",
"Hiroyuki",
""
],
[
"Higuchi",
"Takashi",
""
]
] |
2508.17765
|
Xi Chen
|
Wen-Zhang Wang, Xi Chen, Jin-Ting Li, Dan-Fang Zhang, Wei-Hao Xu, Jia-Yi Wei, Jia-Qi Zhong, Biao Tang, Lin Zhou, Jin Wang, Ming-Sheng Zhan
|
Analysis of the Dick Effect for AI-based Dynamic Gravimeter
|
10 pages,7 figures
| null | null | null |
physics.atom-ph physics.app-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Atom interferometer (AI)-based dynamic gravimeter enable high-precision absolute gravity measurements, crucial for applications in geophysics, navigation, resource exploration, and metrology. Understanding their underlying mechanisms and minimizing measurement noise are essential for enhancing performance. This work investigates the gravity measurement noise in AI-based systems induced by the dead time of the classical accelerometer. Using actual dynamic gravity measurement data, we demonstrate that a dead time of 0.12 s introduces significant gravity measurement noise of 8 mGal. To elucidate the mechanism of this noise, we derive a formula for this noise in frequency domain, identifying high-frequency aliasing as its source. Analysis of the derived expressions indicates that reducing the dead time duration and suppressing the high-frequency noise of the acceleration are effective strategies for mitigating this noise. This work provides significant insights for noise analysis and future scheme design of AI-based dynamic gravimeters.
|
[
{
"version": "v1",
"created": "Mon, 25 Aug 2025 08:02:33 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 08:31:41 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 07:18:55 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Wang",
"Wen-Zhang",
""
],
[
"Chen",
"Xi",
""
],
[
"Li",
"Jin-Ting",
""
],
[
"Zhang",
"Dan-Fang",
""
],
[
"Xu",
"Wei-Hao",
""
],
[
"Wei",
"Jia-Yi",
""
],
[
"Zhong",
"Jia-Qi",
""
],
[
"Tang",
"Biao",
""
],
[
"Zhou",
"Lin",
""
],
[
"Wang",
"Jin",
""
],
[
"Zhan",
"Ming-Sheng",
""
]
] |
2508.17985
|
Abu Shad Ahammed
|
Abu Shad Ahammed, Md Shahi Amran Hossain, Sayeri Mukherjee, Roman Obermaisser, Md. Ziaur Rahman
|
Integration of Computer Vision with Adaptive Control for Autonomous Driving Using ADORE
| null | null | null | null |
cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
Ensuring safety in autonomous driving requires a seamless integration of perception and decision making under uncertain conditions. Although computer vision (CV) models such as YOLO achieve high accuracy in detecting traffic signs and obstacles, their performance degrades in drift scenarios caused by weather variations or unseen objects. This work presents a simulated autonomous driving system that combines a context aware CV model with adaptive control using the ADORE framework. The CARLA simulator was integrated with ADORE via the ROS bridge, allowing real-time communication between perception, decision, and control modules. A simulated test case was designed in both clear and drift weather conditions to demonstrate the robust detection performance of the perception model while ADORE successfully adapted vehicle behavior to speed limits and obstacles with low response latency. The findings highlight the potential of coupling deep learning-based perception with rule-based adaptive decision making to improve automotive safety critical system.
|
[
{
"version": "v1",
"created": "Mon, 25 Aug 2025 12:55:30 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 19:01:33 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Ahammed",
"Abu Shad",
""
],
[
"Hossain",
"Md Shahi Amran",
""
],
[
"Mukherjee",
"Sayeri",
""
],
[
"Obermaisser",
"Roman",
""
],
[
"Rahman",
"Md. Ziaur",
""
]
] |
2508.18222
|
Moustapha Itani
|
Moustapha Itani
|
Exploratory Notes on Symbolic Constraints in Polyhedral Enclosure and Tetrahedral Decomposition in Genus-0 Polyhedra
| null | null | null | null |
cs.CG math.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
I present a coordinate-free, symbolic framework for deciding whether a given set of polygonal faces can form a closed, genus-zero polyhedral surface and for predicting how such a surface could be decomposed into internal tetrahedra. The method uses only discrete incidence variables, such as the number of internal tetrahedra $T$, internal gluing triangles $N_i$, and internal triangulation segments $S_i$, and applies combinatorial feasibility checks before any geometric embedding is attempted. For polyhedra in normal form, I record exact incidence identities linking $V,E,F$ to a flatness parameter $S:=\sum_f(\tmop{deg} f-3)$, and I identify parity-sensitive effects in $E$, $F$, and $S$. The external identities and parity-sensitive bounds hold universally for genus-0 polyhedral graphs. For internal quantities, I prove exact relations $N_i=2T-V+2$ and $T-N_i+S_i=1$ (with $S_i$ taken to be the number of interior edges) and obtain restricted linear ranges within a shell-aligned ladder subclass (SALT), where at most one interior edge is introduced per layer. Consequently, I propose a symbolic workflow that yields rapid pre-checks for structural impossibility, reducing the need for costly geometric validation in computational geometry, graphics, and automated modeling.
|
[
{
"version": "v1",
"created": "Mon, 25 Aug 2025 17:21:13 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Aug 2025 19:44:47 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 13:01:04 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Itani",
"Moustapha",
""
]
] |
2508.18298
|
Gohar Irfan Chaudhry
|
Gohar Irfan Chaudhry, Esha Choukse, Haoran Qiu, \'I\~nigo Goiri, Rodrigo Fonseca, Adam Belay, Ricardo Bianchini
|
Murakkab: Resource-Efficient Agentic Workflow Orchestration in Cloud Platforms
| null | null | null | null |
cs.MA cs.AI cs.SE
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Agentic workflows commonly coordinate multiple models and tools with complex control logic. They are quickly becoming the dominant paradigm for AI applications. However, serving them remains inefficient with today's frameworks. The key problem is that they expose workflows as opaque sequences of model and tool calls that tightly couple agent logic with model and hardware choices. Often, these workflow components are fragmented across different entities, preventing systems from reasoning about trade-offs across accuracy, latency, energy, and cost. This leads to resource waste and degraded service-level objectives (SLOs).
We present Murakkab, a resource-efficient serving system for agentic workflows. Murakkab introduces a declarative abstraction that decouples workflow specification from execution configuration. A profile-guided optimizer and adaptive runtime jointly manage the full stack: orchestrating workflow components, mapping them to models and hardware, and dynamically reconfiguring execution to satisfy user-defined SLOs. By exposing the internal structure of agentic workflows, Murakkab enables cross-layer optimization that existing frameworks and cloud schedulers cannot achieve.
Our evaluation on diverse workflows shows that Murakkab reduces GPU usage by up to 2.8$\times$, energy consumption by 3.7$\times$, and cost by 4.3$\times$ while maintaining SLOs.
|
[
{
"version": "v1",
"created": "Fri, 22 Aug 2025 17:41:27 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 16:28:25 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Chaudhry",
"Gohar Irfan",
""
],
[
"Choukse",
"Esha",
""
],
[
"Qiu",
"Haoran",
""
],
[
"Goiri",
"Íñigo",
""
],
[
"Fonseca",
"Rodrigo",
""
],
[
"Belay",
"Adam",
""
],
[
"Bianchini",
"Ricardo",
""
]
] |
2508.18310
|
Seungju Lee
|
Seung Ju Lee
|
Weight-Shifting Operators of Hypergeometric Type for Maass Forms
|
v4: Corrected a calculation error in the formula for the growth exponent alpha_f (Lemma 5.1). This correction affects the combined exponent alpha_K and the final expression for the region of absolute convergence, R_abs (Theorem 4.3). The abstract and introduction have been revised accordingly
| null | null | null |
math.NT math.RT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper constructs a weight-shifting integral operator, U_{k,t}, that maps the space of weight-k Maass forms to the space of weight-t functions for the full modular group SL(2, Z). The operator is defined as an integral transform whose kernel, K_{k,t,q}, is the product of a specially designed covariant factor P and a modular invariant function F. The central hypothesis of our method is that the kernel itself must be an eigenfunction of the weight-t hyperbolic Laplacian Delta_t with respect to its first variable. We show that this spectral condition reduces the problem of determining the invariant function F to a second-order ordinary differential equation. The principal result of this work is the identification of this ODE as a Papperitz-Riemann equation, which is then transformed into the standard Gauss Hypergeometric Differential Equation. We provide an explicit determination of the hypergeometric parameters (a,b,c) in terms of the initial automorphic data. Furthermore, we conduct a detailed analysis of the operator's convergence, rigorously establishing the region of absolute convergence in the parameter space by examining the asymptotic behavior of the kernel. The emergence of the hypergeometric structure suggests a deep connection between these weight-shifting operators and the theory of hypergeometric motives in arithmetic geometry, a principle supported by the recovery of the periods of the Legendre family of elliptic curves in a specific case.
|
[
{
"version": "v1",
"created": "Sat, 23 Aug 2025 11:55:02 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 22:30:02 GMT"
},
{
"version": "v3",
"created": "Sun, 31 Aug 2025 09:14:04 GMT"
},
{
"version": "v4",
"created": "Wed, 03 Sep 2025 16:25:04 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lee",
"Seung Ju",
""
]
] |
2508.18453
|
Yaser Baseri
|
Yaser Baseri, Abdelhakim Senhaji Hafid, and Dimitrios Makrakis and Hamidreza Fereidouni
|
Privacy-Preserving Federated Learning Framework for Risk-Based Adaptive Authentication
| null | null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Balancing robust security with strong privacy guarantees is critical for Risk-Based Adaptive Authentication (RBA), particularly in decentralized settings. Federated Learning (FL) offers a promising solution by enabling collaborative risk assessment without centralizing user data. However, existing FL approaches struggle with Non-Independent and Identically Distributed (Non-IID) user features, resulting in biased, unstable, and poorly generalized global models. This paper introduces FL-RBA2, a novel Federated Learning framework for Risk-Based Adaptive Authentication that addresses Non-IID challenges through a mathematically grounded similarity transformation. By converting heterogeneous user features (including behavioral, biometric, contextual, interaction-based, and knowledge-based modalities) into IID similarity vectors, FL-RBA2 supports unbiased aggregation and personalized risk modeling across distributed clients. The framework mitigates cold-start limitations via clustering-based risk labeling, incorporates Differential Privacy (DP) to safeguard sensitive information, and employs Message Authentication Codes (MACs) to ensure model integrity and authenticity. Federated updates are securely aggregated into a global model, achieving strong balance between user privacy, scalability, and adaptive authentication robustness. Rigorous game-based security proofs in the Random Oracle Model formally establish privacy, correctness, and adaptive security guarantees. Extensive experiments on keystroke, mouse, and contextual datasets validate FL-RBA2's effectiveness in high-risk user detection and its resilience to model inversion and inference attacks, even under strong DP constraints.
|
[
{
"version": "v1",
"created": "Mon, 25 Aug 2025 20:02:07 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 00:17:00 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Baseri",
"Yaser",
""
],
[
"Hafid",
"Abdelhakim Senhaji",
""
],
[
"Makrakis",
"Dimitrios",
""
],
[
"Fereidouni",
"Hamidreza",
""
]
] |
2508.18721
|
Yunrui Pei
|
Yunrui Pei, Hongshu Wang, Wenjie Zhang, Yun Lin, Weiyu Kong, Jin song Dong
|
LLM as an Execution Estimator: Recovering Missing Dependency for Practical Time-travelling Debugging
| null | null | null | null |
cs.SE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Determining the dynamic data dependency of a step that reads a variable $v$ is challenging. It typically requires either exhaustive instrumentation, which becomes prohibitively expensive when $v$ is defined within library calls, or repeated executions, which are impractical for non-deterministic programs. In this work, we propose RecovSlicing for computing dynamic data dependency in a single run, with only partial instrumentation. We explore the intuition that LLM can potentially infer program dynamics based on a partially recorded trace and relevant code as its context. Given (1) a partially recorded trace of a program $P$ and (2) the slicing criteria consisting of a query step $s$ and a query variable $v$ read by $s$, RecovSlicing computes the runtime definition of $v$ on the trace by estimating the miss-recorded execution of $P$. In this work, we allow the user to specify implicit query variable. Technically, built upon non-deterministic LLM, we address the challenges of (1) precise recovery of runtime variable value and structure from the recorded execution and (2) aligning the memory address of recovered variables and the recorded variables for definition analysis. We evaluate RecovSlicing on 8300 data dependencies across three slicing benchmarks, comparing it with Slicer4J, ND-Slicer, LLM Slicer, and re-execution Slicer. RecovSlicing achieves significantly higher accuracy (80.3%, 91.1%, 98.3%) and recall (up to 98.3%) than the best baseline (accuracy: 39.0%, 82.0%, 59.9%; recall: 53.4%, 79.1%, 87.1%). Integrated into a dual-slicing regression bug localizer, it identifies 16% more regressions.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 06:38:38 GMT"
},
{
"version": "v2",
"created": "Wed, 27 Aug 2025 03:44:39 GMT"
},
{
"version": "v3",
"created": "Thu, 28 Aug 2025 03:04:37 GMT"
},
{
"version": "v4",
"created": "Fri, 29 Aug 2025 01:09:40 GMT"
},
{
"version": "v5",
"created": "Wed, 03 Sep 2025 14:34:30 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Pei",
"Yunrui",
""
],
[
"Wang",
"Hongshu",
""
],
[
"Zhang",
"Wenjie",
""
],
[
"Lin",
"Yun",
""
],
[
"Kong",
"Weiyu",
""
],
[
"Dong",
"Jin song",
""
]
] |
2508.18727
|
Shantanu Semwal
|
Shantanu Semwal, Yi Xia, Chris Wolverton, Koushik Pal
|
Origin of Glass-like Thermal Conductivity in Crystalline TlAgTe
| null | null | null | null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Ordered crystalline compounds exhibiting ultralow and glass-like thermal conductivity are both fundamentally and technologically important, where phonon quasi-particles dominate their heat transport. Understanding the microscopic mechanisms that govern such unusual transport behavior is necessary to unravel the complex interplay of crystal structure, phonons, and collective excitations of these quasi-particles. Here, we use state-of-the-art first-principles calculations based on quantum density functional theory to investigate the origin of experimentally measured unusually low and glassy thermal conductivity in semiconducting TlAgTe. Utilizing a unifying framework of anharmonic lattice dynamics theory that combine phonon self-energy induced frequency renormalization, particle-like Peierls ($\kappa_{l}^{P}$) and wave-like coherent ($\kappa_{l}^{C}$) thermal transport contributions including three and four-phonon scattering channels, we successfully explain the experimental results both in terms of magnitude and temperature dependence. Our analysis reveals that TlAgTe exhibits several localized phonon modes arising from concerted rattling-like vibration of Tl atoms, which show strong temperature dependence and enhanced four-phonon scattering rates that are dominated by Umklapp processes, suppressing $\kappa_{l}^{P}$ to ultralow values. The ensuing strong anharmonicity induced by local structural distortions, lone-pair electrons, and rattling-like vibrations of the heavy cations lead to a transition from particle-like behavior to wave-like tunneling characteristics of the phonon modes above 40 cm$^{-1}$, contributing significantly to $\kappa_{l}^{C}$ which increases with temperature. Our analysis uncovers important structure-property relationship, which may be used in designing of novel materials with tunable thermal conductivity.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 06:52:10 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 19:16:40 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Semwal",
"Shantanu",
""
],
[
"Xia",
"Yi",
""
],
[
"Wolverton",
"Chris",
""
],
[
"Pal",
"Koushik",
""
]
] |
2508.18757
|
Chuan Hung Chen
|
Chuan-Hung Chen
|
Structure of leptonic Yukawa couplings in the Zee model
|
11 pages, references added
| null | null | null |
hep-ph hep-ex
|
http://creativecommons.org/publicdomain/zero/1.0/
|
The radiative neutrino mass matrix $m^\nu$ in the Zee model depends on leptonic Yukawa couplings $F$ to a singlet scalar and $Y^\ell$ to a new Higgs doublet. Leveraging the skew-symmetric structure of $F$, we derive a unique identity linking $F$ and $m^\nu$ that is explicitly independent of $Y^\ell$. This relation implies that five entries of $Y^\ell$ can, in principle, be determined directly from $m^\nu$ and $F$, while the remaining four can be selected based on phenomenological assumptions. As an illustration, we apply this framework to the two-zero texture $B2$, highlighting its enhancement of the muon $g-2$.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 07:35:18 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 15:24:04 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Chen",
"Chuan-Hung",
""
]
] |
2508.19200
|
Xinran Zhao
|
Xinran Zhao, Boyuan Zheng, Chenglei Si, Haofei Yu, Ken Liu, Runlong Zhou, Ruochen Li, Tong Chen, Xiang Li, Yiming Zhang, Tongshuang Wu
|
The Ramon Llull's Thinking Machine for Automated Ideation
|
21 pages, 3 figures
| null | null | null |
cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
This paper revisits Ramon Llull's Ars combinatoria - a medieval framework for generating knowledge through symbolic recombination - as a conceptual foundation for building a modern Llull's thinking machine for research ideation. Our approach defines three compositional axes: Theme (e.g., efficiency, adaptivity), Domain (e.g., question answering, machine translation), and Method (e.g., adversarial training, linear attention). These elements represent high-level abstractions common in scientific work - motivations, problem settings, and technical approaches - and serve as building blocks for LLM-driven exploration. We mine elements from human experts or conference papers and show that prompting LLMs with curated combinations produces research ideas that are diverse, relevant, and grounded in current literature. This modern thinking machine offers a lightweight, interpretable tool for augmenting scientific creativity and suggests a path toward collaborative ideation between humans and AI.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 17:03:43 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Aug 2025 17:29:36 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 16:14:12 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Zhao",
"Xinran",
""
],
[
"Zheng",
"Boyuan",
""
],
[
"Si",
"Chenglei",
""
],
[
"Yu",
"Haofei",
""
],
[
"Liu",
"Ken",
""
],
[
"Zhou",
"Runlong",
""
],
[
"Li",
"Ruochen",
""
],
[
"Chen",
"Tong",
""
],
[
"Li",
"Xiang",
""
],
[
"Zhang",
"Yiming",
""
],
[
"Wu",
"Tongshuang",
""
]
] |
2508.19352
|
Adarsh Jamadandi
|
Adarsh Jamadandi, Jing Xu, Adam Dziedzic, Franziska Boenisch
|
Memorization in Graph Neural Networks
|
Version3, updated affiliation
| null | null | null |
cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Deep neural networks (DNNs) have been shown to memorize their training data, yet similar analyses for graph neural networks (GNNs) remain largely under-explored. We introduce NCMemo (Node Classification Memorization), the first framework to quantify label memorization in semi-supervised node classification. We first establish an inverse relationship between memorization and graph homophily, i.e., the property that connected nodes share similar labels/features. We find that lower homophily significantly increases memorization, indicating that GNNs rely on memorization to learn less homophilic graphs. Secondly, we analyze GNN training dynamics. We find that the increased memorization in low homophily graphs is tightly coupled to the GNNs' implicit bias on using graph structure during learning. In low homophily regimes, this structure is less informative, hence inducing memorization of the node labels to minimize training loss. Finally, we show that nodes with higher label inconsistency in their feature-space neighborhood are significantly more prone to memorization. Building on our insights into the link between graph homophily and memorization, we investigate graph rewiring as a means to mitigate memorization. Our results demonstrate that this approach effectively reduces memorization without compromising model performance. Moreover, we show that it lowers the privacy risk for previously memorized data points in practice. Thus, our work not only advances understanding of GNN learning but also supports more privacy-preserving GNN deployment.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 18:25:54 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Aug 2025 19:24:43 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 08:38:08 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Jamadandi",
"Adarsh",
""
],
[
"Xu",
"Jing",
""
],
[
"Dziedzic",
"Adam",
""
],
[
"Boenisch",
"Franziska",
""
]
] |
2508.19362
|
Pranav Rajbhandari
|
Florian Frick and Pranav Rajbhandari
|
Geodesic complexity of the octahedron, and an algorithm for cut loci on convex polyhedra
|
44 pages, 26 figures
| null | null | null |
math.MG cs.CG
|
http://creativecommons.org/licenses/by/4.0/
|
The geodesic complexity of a length space $X$ quantifies the required number of case distinctions to continuously choose a shortest path connecting any given start and end point. We prove a local lower bound for the geodesic complexity of $X$ obtained by embedding simplices into $X\times X$. We additionally create and prove correctness of an algorithm to find cut loci on surfaces of convex polyhedra, as the structure of a space's cut loci is related to its geodesic complexity. We use these techniques to prove the geodesic complexity of the octahedron is four. Our method is inspired by earlier work of Recio-Mitter and Davis, and thus recovers their results on the geodesic complexity of the $n$-torus and the tetrahedron, respectively.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 18:40:58 GMT"
},
{
"version": "v2",
"created": "Fri, 29 Aug 2025 16:48:00 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 04:48:12 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Frick",
"Florian",
""
],
[
"Rajbhandari",
"Pranav",
""
]
] |
2508.19444
|
Suhala Rabab Saba
|
Suhala Rabab Saba, Sagar Dasgupta, Mizanur Rahman, Nathan Huynh, Li Zhao, Mehmet C. Vuran, Qiang Liu, Eren Erman Ozguven
|
Infrastructure-enabled risk assessment of hazardous road conditions on rural roads during inclement weather
|
18 pages, 5 figures, 5 tables
| null | null | null |
cs.CE
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Rural roadways often expose Commercial Motor Vehicle (CMV) drivers to hazardous conditions, such as heavy fog, rain, snow, black ice, and flash floods, many of which remain unreported in real time. This lack of timely information, coupled with limited infrastructure in rural areas, significantly increases the risk of crashes. Although various sensing technologies exist to monitor individual hazards like low visibility or surface friction, they rarely assess the combined driving risk posed by multiple simultaneous hazards, nor do they provide actionable recommendations such as safe advisory speeds. To address this critical gap, in this study, we present a roadway hazard risk assessment framework that provides an approach to quantify the probability and severity of crash occurrences due to specific roadway hazards. To evaluate this framework, we presented a case study by constructing a synthetic "year-long" dataset that encompasses every possible pairing of road surface and visibility conditions. Our analysis confirms that the combined ProbabilitySeverity scoring yields a coherent, stepwise risk profile across all hazard scenarios. These results validate the practicality of our risk assessment approach and provide a foundation for deploying graduated safety measures in real-world roadway operations.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 21:32:40 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 22:02:27 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Saba",
"Suhala Rabab",
""
],
[
"Dasgupta",
"Sagar",
""
],
[
"Rahman",
"Mizanur",
""
],
[
"Huynh",
"Nathan",
""
],
[
"Zhao",
"Li",
""
],
[
"Vuran",
"Mehmet C.",
""
],
[
"Liu",
"Qiang",
""
],
[
"Ozguven",
"Eren Erman",
""
]
] |
2508.19493
|
Jungang Li
|
Zhixin Lin, Jungang Li, Shidong Pan, Yibo Shi, Yue Yao, Dongliang Xu
|
Mind the Third Eye! Benchmarking Privacy Awareness in MLLM-powered Smartphone Agents
| null | null | null | null |
cs.CR cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Smartphones bring significant convenience to users but also enable devices to extensively record various types of personal information. Existing smartphone agents powered by Multimodal Large Language Models (MLLMs) have achieved remarkable performance in automating different tasks. However, as the cost, these agents are granted substantial access to sensitive users' personal information during this operation. To gain a thorough understanding of the privacy awareness of these agents, we present the first large-scale benchmark encompassing 7,138 scenarios to the best of our knowledge. In addition, for privacy context in scenarios, we annotate its type (e.g., Account Credentials), sensitivity level, and location. We then carefully benchmark seven available mainstream smartphone agents. Our results demonstrate that almost all benchmarked agents show unsatisfying privacy awareness (RA), with performance remaining below 60% even with explicit hints. Overall, closed-source agents show better privacy ability than open-source ones, and Gemini 2.0-flash achieves the best, achieving an RA of 67%. We also find that the agents' privacy detection capability is highly related to scenario sensitivity level, i.e., the scenario with a higher sensitivity level is typically more identifiable. We hope the findings enlighten the research community to rethink the unbalanced utility-privacy tradeoff about smartphone agents. Our code and benchmark are available at https://zhixin-l.github.io/SAPA-Bench.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 00:41:28 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:48:34 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lin",
"Zhixin",
""
],
[
"Li",
"Jungang",
""
],
[
"Pan",
"Shidong",
""
],
[
"Shi",
"Yibo",
""
],
[
"Yao",
"Yue",
""
],
[
"Xu",
"Dongliang",
""
]
] |
2508.19599
|
Jonathan De Brusse
|
Jonathan de Brusse, Jamal Daafouz, Mathieu Granzotto, Romain Postoyan and Dragan Nesic
|
Discounted LQR: stabilizing (near-)optimal state-feedback laws
| null | null | null | null |
math.OC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study deterministic, discrete linear time-invariant systems with infinite-horizon discounted quadratic cost. It is well-known that standard stabilizability and detectability properties are not enough in general to conclude stability properties for the system in closed-loop with the optimal controller when the discount factor is small. In this context, we first review some of the stability conditions based on the optimal value function found in the learning and control literature and highlight their conservatism. We then propose novel (necessary and) sufficient conditions, still based on the optimal value function, under which stability of the origin for the optimal closed-loop system is guaranteed. Afterwards, we focus on the scenario where the optimal feedback law is not stabilizing because of the discount factor and the goal is to design an alternative stabilizing near-optimal static state-feedback law. We present both linear matrix inequality-based conditions and a variant of policy iteration to construct such stabilizing near-optimal controllers. The methods are illustrated via numerical examples.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 06:20:04 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 06:53:43 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"de Brusse",
"Jonathan",
""
],
[
"Daafouz",
"Jamal",
""
],
[
"Granzotto",
"Mathieu",
""
],
[
"Postoyan",
"Romain",
""
],
[
"Nesic",
"Dragan",
""
]
] |
2508.19660
|
Konstantinos Balaskas
|
Vojtech Mrazek, Konstantinos Balaskas, Paula Carolina Lozano Duarte, Zdenek Vasicek, Mehdi B. Tahoori, Georgios Zervakis
|
Arbitrary Precision Printed Ternary Neural Networks with Holistic Evolutionary Approximation
|
Accepted for publication at IEEE Transactions on Circuits and Systems for Artificial Intelligence (TCASAI)
| null |
10.1109/TCASAI.2025.3604384
| null |
eess.SP cs.AI cs.NE
|
http://creativecommons.org/licenses/by/4.0/
|
Printed electronics offer a promising alternative for applications beyond silicon-based systems, requiring properties like flexibility, stretchability, conformality, and ultra-low fabrication costs. Despite the large feature sizes in printed electronics, printed neural networks have attracted attention for meeting target application requirements, though realizing complex circuits remains challenging. This work bridges the gap between classification accuracy and area efficiency in printed neural networks, covering the entire processing-near-sensor system design and co-optimization from the analog-to-digital interface-a major area and power bottleneck-to the digital classifier. We propose an automated framework for designing printed Ternary Neural Networks with arbitrary input precision, utilizing multi-objective optimization and holistic approximation. Our circuits outperform existing approximate printed neural networks by 17x in area and 59x in power on average, being the first to enable printed-battery-powered operation with under 5% accuracy loss while accounting for analog-to-digital interfacing costs.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 08:19:22 GMT"
},
{
"version": "v2",
"created": "Fri, 29 Aug 2025 13:07:04 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 07:30:53 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Mrazek",
"Vojtech",
""
],
[
"Balaskas",
"Konstantinos",
""
],
[
"Duarte",
"Paula Carolina Lozano",
""
],
[
"Vasicek",
"Zdenek",
""
],
[
"Tahoori",
"Mehdi B.",
""
],
[
"Zervakis",
"Georgios",
""
]
] |
2508.19805
|
Shota Naito
|
Shota Naito, Tsukasa Ninomiya and Koichi Wada
|
Separation of Three or More Autonomous Mobile Models under Hierarchical Schedulers
| null | null | null | null |
cs.DC cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
Understanding the computational power of mobile robot systems is a fundamental challenge in distributed computing. While prior work has focused on pairwise separations between models, we explore how robot capabilities, light observability, and scheduler synchrony interact in more complex ways.
We first show that the Exponential Times Expansion (ETE) problem is solvable only in the strongest model -- fully-synchronous robots with full mutual lights ($\mathcal{LUMT}^F$). We then introduce the Hexagonal Edge Traversal (HET) and TAR(d)* problems to demonstrate how internal memory and lights interact with synchrony: under weak synchrony, internal memory alone is insufficient, while full synchrony can substitute for both lights and memory.
In the asynchronous setting, we classify problems such as LP-MLCv, VEC, and ZCC to show fine-grained separations between $\mathcal{FSTA}$ and $\mathcal{FCOM}$ robots. We also analyze Vertex Traversal Rendezvous (VTR) and Leave Place Convergence (LP-Cv), illustrating the limitations of internal memory in symmetric settings.
These results extend the known separation map of 14 canonical robot models, revealing structural phenomena only visible through higher-order comparisons. Our work provides new impossibility criteria and deepens the understanding of how observability, memory, and synchrony collectively shape the computational power of mobile robots.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 11:47:29 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Naito",
"Shota",
""
],
[
"Ninomiya",
"Tsukasa",
""
],
[
"Wada",
"Koichi",
""
]
] |
2508.19828
|
Sikuan Yan
|
Sikuan Yan, Xiufeng Yang, Zuchao Huang, Ercong Nie, Zifeng Ding, Zonggen Li, Xiaowen Ma, Hinrich Sch\"utze, Volker Tresp, Yunpu Ma
|
Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning
|
work in progress
| null | null | null |
cs.CL cs.MA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large Language Models (LLMs) have demonstrated impressive capabilities across a wide range of NLP tasks, but they remain fundamentally stateless, constrained by limited context windows that hinder long-horizon reasoning. Recent efforts to address this limitation often augment LLMs with an external memory bank, yet most existing pipelines are static and heuristic-driven, lacking any learned mechanism for deciding what to store, update, or retrieve. We present Memory-R1, a reinforcement learning (RL) framework that equips LLMs with the ability to actively manage and utilize external memory through two specialized agents: a Memory Manager that learns to perform structured memory operations, including adding, updating, deleting, or taking no operation on memory entries; and an Answer Agent that selects the most relevant entries and reasons over them to produce an answer. Both agents are fine-tuned with outcome-driven RL (PPO and GRPO), enabling adaptive memory management and utilization with minimal supervision. With as few as 152 question-answer pairs and a corresponding temporal memory bank for training, Memory-R1 outperforms the strongest existing baseline and demonstrates strong generalization across diverse question types and LLM backbones. Beyond presenting an effective approach, this work provides insights into how RL can unlock more agentic, memory-aware behavior in LLMs, pointing toward richer, more persistent reasoning systems.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 12:26:55 GMT"
},
{
"version": "v2",
"created": "Fri, 29 Aug 2025 04:52:28 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 09:33:30 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Yan",
"Sikuan",
""
],
[
"Yang",
"Xiufeng",
""
],
[
"Huang",
"Zuchao",
""
],
[
"Nie",
"Ercong",
""
],
[
"Ding",
"Zifeng",
""
],
[
"Li",
"Zonggen",
""
],
[
"Ma",
"Xiaowen",
""
],
[
"Schütze",
"Hinrich",
""
],
[
"Tresp",
"Volker",
""
],
[
"Ma",
"Yunpu",
""
]
] |
2508.19855
|
Junnan Dong
|
Junnan Dong, Siyu An, Yifei Yu, Qian-Wen Zhang, Linhao Luo, Xiao Huang, Yunsheng Wu, Di Yin, Xing Sun
|
Youtu-GraphRAG: Vertically Unified Agents for Graph Retrieval-Augmented Complex Reasoning
|
19 pages, 7 figures, 6 tables
| null | null | null |
cs.IR
|
http://creativecommons.org/licenses/by/4.0/
|
Graph retrieval-augmented generation (GraphRAG) has effectively enhanced large language models in complex reasoning by organizing fragmented knowledge into explicitly structured graphs. Prior efforts have been made to improve either graph construction or graph retrieval in isolation, yielding suboptimal performance, especially when domain shifts occur. In this paper, we propose a vertically unified agentic paradigm, Youtu-GraphRAG, to jointly connect the entire framework as an intricate integration. Specifically, (i) a seed graph schema is introduced to bound the automatic extraction agent with targeted entity types, relations and attribute types, also continuously expanded for scalability over unseen domains; (ii) To obtain higher-level knowledge upon the schema, we develop novel dually-perceived community detection, fusing structural topology with subgraph semantics for comprehensive knowledge organization. This naturally yields a hierarchical knowledge tree that supports both top-down filtering and bottom-up reasoning with community summaries; (iii) An agentic retriever is designed to interpret the same graph schema to transform complex queries into tractable and parallel sub-queries. It iteratively performs reflection for more advanced reasoning; (iv) To alleviate the knowledge leaking problem in pre-trained LLM, we propose a tailored anonymous dataset and a novel 'Anonymity Reversion' task that deeply measures the real performance of the GraphRAG frameworks. Extensive experiments across six challenging benchmarks demonstrate the robustness of Youtu-GraphRAG, remarkably moving the Pareto frontier with up to 90.71% saving of token costs and 16.62% higher accuracy over state-of-the-art baselines. The results indicate our adaptability, allowing seamless domain transfer with minimal intervention on schema.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 13:13:20 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 08:17:51 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 02:56:22 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Dong",
"Junnan",
""
],
[
"An",
"Siyu",
""
],
[
"Yu",
"Yifei",
""
],
[
"Zhang",
"Qian-Wen",
""
],
[
"Luo",
"Linhao",
""
],
[
"Huang",
"Xiao",
""
],
[
"Wu",
"Yunsheng",
""
],
[
"Yin",
"Di",
""
],
[
"Sun",
"Xing",
""
]
] |
2508.19897
|
Luca Ambrogioni
|
Luca Ambrogioni
|
The Information Dynamics of Generative Diffusion
| null | null | null | null |
stat.ML cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Generative diffusion models have emerged as a powerful class of models in machine learning, yet a unified theoretical understanding of their operation is still developing. This perspective paper provides an integrated perspective on generative diffusion by connecting their dynamic, information-theoretic, and thermodynamic properties under a unified mathematical framework. We demonstrate that the rate of conditional entropy production during generation (i.e. the generative bandwidth) is directly governed by the expected divergence of the score function's vector field. This divergence, in turn, is linked to the branching of trajectories and generative bifurcations, which we characterize as symmetry-breaking phase transitions in the energy landscape. This synthesis offers a powerful insight: the process of generation is fundamentally driven by the controlled, noise-induced breaking of (approximate) symmetries, where peaks in information transfer correspond to critical transitions between possible outcomes. The score function acts as a dynamic non-linear filter that regulates the bandwidth of the noise by suppressing fluctuations that are incompatible with the data.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 13:53:56 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 10:38:15 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Ambrogioni",
"Luca",
""
]
] |
2508.20282
|
Hyejun Jeong
|
Hyejun Jeong, Mohammadreza Teymoorianfard, Abhinav Kumar, Amir Houmansadr, Eugene Bagdasarian
|
Network-Level Prompt and Trait Leakage in Local Research Agents
|
Code available at https://github.com/umass-aisec/wra
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We show that Web and Research Agents (WRAs) -- language model-based systems that investigate complex topics on the Internet -- are vulnerable to inference attacks by passive network adversaries such as ISPs. These agents could be deployed locally by organizations and individuals for privacy, legal, or financial purposes. Unlike sporadic web browsing by humans, WRAs visit $70{-}140$ domains with distinguishable timing correlations, enabling unique fingerprinting attacks.
Specifically, we demonstrate a novel prompt and user trait leakage attack against WRAs that only leverages their network-level metadata (i.e., visited IP addresses and their timings). We start by building a new dataset of WRA traces based on user search queries and queries generated by synthetic personas. We define a behavioral metric (called OBELS) to comprehensively assess similarity between original and inferred prompts, showing that our attack recovers over 73% of the functional and domain knowledge of user prompts. Extending to a multi-session setting, we recover up to 19 of 32 latent traits with high accuracy. Our attack remains effective under partial observability and noisy conditions. Finally, we discuss mitigation strategies that constrain domain diversity or obfuscate traces, showing negligible utility impact while reducing attack effectiveness by an average of 29%.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 21:24:10 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 03:33:59 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Jeong",
"Hyejun",
""
],
[
"Teymoorianfard",
"Mohammadreza",
""
],
[
"Kumar",
"Abhinav",
""
],
[
"Houmansadr",
"Amir",
""
],
[
"Bagdasarian",
"Eugene",
""
]
] |
2508.20553
|
Alexander Gr\"afe
|
Alexander Gr\"afe, Joram Eickhoff, Marco Zimmerling, Sebastian Trimpe
|
DMPC-Swarm: Distributed Model Predictive Control on Nano UAV Swarms
| null | null | null | null |
eess.SY cs.SY
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Swarms of unmanned aerial vehicles (UAVs) are increasingly becoming vital to our society, undertaking tasks such as search and rescue, surveillance and delivery. A special variant of Distributed Model Predictive Control (DMPC) has emerged as a promising approach for the safe management of these swarms by combining the scalability of distributed computation with dynamic swarm motion control. In this DMPC method, multiple agents solve local optimization problems with coupled anti-collision constraints, periodically exchanging their solutions. Despite its potential, existing methodologies using this DMPC variant have yet to be deployed on distributed hardware that fully utilize true distributed computation and wireless communication. This is primarily due to the lack of a communication system tailored to meet the unique requirements of mobile swarms and an architecture that supports distributed computation while adhering to the payload constraints of UAVs. We present DMPC-SWARM, a new swarm control methodology that integrates an efficient, stateless low-power wireless communication protocol with a novel DMPC algorithm that provably avoids UAV collisions even under message loss. By utilizing event-triggered and distributed off-board computing, DMPC-SWARM supports nano UAVs, allowing them to benefit from additional computational resources while retaining scalability and fault tolerance. In a detailed theoretical analysis, we prove that DMPC-SWARM guarantees collision avoidance under realistic conditions, including communication delays and message loss. Finally, we present DMPC-SWARM's implementation on a swarm of up to 16 nano-quadcopters, demonstrating the first realization of these DMPC variants with computation distributed on multiple physical devices interconnected by a real wireless mesh networks. A video showcasing DMPC-SWARM is available at http://tiny.cc/DMPCSwarm.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 08:45:45 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 13:38:23 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Gräfe",
"Alexander",
""
],
[
"Eickhoff",
"Joram",
""
],
[
"Zimmerling",
"Marco",
""
],
[
"Trimpe",
"Sebastian",
""
]
] |
2508.20748
|
Jun Xie
|
Jun Xie, Yuan-Hua Ni, Yiqin Yang, Bo Xu
|
An Efficient Data-Driven Framework for Linear Quadratic Output Feedback Control
| null | null | null | null |
math.OC cs.SY eess.SY
|
http://creativecommons.org/licenses/by/4.0/
|
Linear quadratic regulator with unmeasurable states and unknown system matrix parameters better aligns with practical scenarios. However, for this problem, balancing the optimality of the resulting controller and the leniency of the algorithm's feasibility conditions remains a non-trivial challenge, as no well-established general method has yet been developed to address this trade-off. To address this gap, this study first develops a comprehensive theoretical framework for state parameterization that equivalently substitutes for unknown states. By analyzing the controllability of consistent systems satisfied by substitute states, this framework quantifies the capability of substitute state data matrices to parameterize unknown closed-loop systems and output feedback controllers, thereby constructing a modified state parameterization form that meets the complete data parameterization condition of Willems' Fundamental Lemma. Leveraging this framework, this study proposes efficient model-free off-policy policy iteration and value iteration algorithms with theoretical guarantees to solve for the optimal output feedback controller. Compared with existing studies, particularly for multi-output problems where existing model-free reinforcement learning algorithms may fail, the proposed method removes redundant information in substitute states and the additional full row rank condition on regression matrices, thereby ensuring the solution of optimal output feedback controllers equivalent to optimal state feedback controllers for multi-output systems. Furthermore, this study pioneers a comprehensive and highly scalable theoretical analysis of state parameterization from a data-driven viewpoint, and the proposed algorithms exhibit significant advantages in implementation conditions, data demand, unknown handling, and convergence speed.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 13:07:59 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 11:28:58 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Xie",
"Jun",
""
],
[
"Ni",
"Yuan-Hua",
""
],
[
"Yang",
"Yiqin",
""
],
[
"Xu",
"Bo",
""
]
] |
2508.20757
|
Yuanhao Ding
|
Yuanhao Ding and Esteban Garces Arias and Meimingwei Li and Julian Rodemann and Matthias A{\ss}enmacher and Danlu Chen and Gaojuan Fan and Christian Heumann and Chongsheng Zhang
|
GUARD: Glocal Uncertainty-Aware Robust Decoding for Effective and Efficient Open-Ended Text Generation
|
Accepted at Findings of the Association for Computational Linguistics: EMNLP 2025
| null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Open-ended text generation faces a critical challenge: balancing coherence with diversity in LLM outputs. While contrastive search-based decoding strategies have emerged to address this trade-off, their practical utility is often limited by hyperparameter dependence and high computational costs. We introduce GUARD, a self-adaptive decoding method that effectively balances these competing objectives through a novel "Glocal" uncertainty-driven framework. GUARD combines global entropy estimates with local entropy deviations to integrate both long-term and short-term uncertainty signals. We demonstrate that our proposed global entropy formulation effectively mitigates abrupt variations in uncertainty, such as sudden overconfidence or high entropy spikes, and provides theoretical guarantees of unbiasedness and consistency. To reduce computational overhead, we incorporate a simple yet effective token-count-based penalty into GUARD. Experimental results demonstrate that GUARD achieves a good balance between text diversity and coherence, while exhibiting substantial improvements in generation speed. In a more nuanced comparison study across different dimensions of text quality, both human and LLM evaluators validated its remarkable performance. Our code is available at https://github.com/YecanLee/GUARD.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 13:14:20 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:21:02 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Ding",
"Yuanhao",
""
],
[
"Arias",
"Esteban Garces",
""
],
[
"Li",
"Meimingwei",
""
],
[
"Rodemann",
"Julian",
""
],
[
"Aßenmacher",
"Matthias",
""
],
[
"Chen",
"Danlu",
""
],
[
"Fan",
"Gaojuan",
""
],
[
"Heumann",
"Christian",
""
],
[
"Zhang",
"Chongsheng",
""
]
] |
2508.21161
|
Lucas Marques Fagundes Silva
|
L. Marques Fagundes Silva, R. C. Souza Pimenta, M. H. Magiotto, R. M. Gomes, E. I. Duzzioni, R. Medeiros de Ara\'ujo, P. H. Souto Ribeiro
|
Optical Integration With Heralded Single Photons
| null | null | null | null |
quant-ph physics.optics
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this work, we demonstrate optical integration using heralded single photons and explore the influence of spatial correlations between photons on this process. Specifically, we experimentally harness the transverse spatial degrees of freedom of light within an optical processing framework based on heralded single photons. The integration is performed over binary phase patterns encoded via a phase-only spatial light modulator, with polarization serving as an auxiliary degree of freedom. Our findings reveal a distinct contrast in how spatial correlations affect image analysis: spatially uncorrelated photons are more effective at capturing the global features of an image encoded in the modulator, whereas spatially correlated photons exhibit enhanced sensitivity to local image details. Importantly, the optical integration scheme presented here bears a strong conceptual and operational resemblance to the DQC1 (Deterministic Quantum Computation with One Qubit) model. This connection underscores the potential of our approach for quantum-enhanced information processing, even in regimes where entanglement is minimal or absent.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 18:52:43 GMT"
},
{
"version": "v2",
"created": "Mon, 01 Sep 2025 01:37:09 GMT"
},
{
"version": "v3",
"created": "Wed, 03 Sep 2025 00:23:44 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Silva",
"L. Marques Fagundes",
""
],
[
"Pimenta",
"R. C. Souza",
""
],
[
"Magiotto",
"M. H.",
""
],
[
"Gomes",
"R. M.",
""
],
[
"Duzzioni",
"E. I.",
""
],
[
"de Araújo",
"R. Medeiros",
""
],
[
"Ribeiro",
"P. H. Souto",
""
]
] |
2508.21236
|
Malte L\"uken
|
Malte L\"uken, Javier Garcia-Bernardo, Sreeparna Deb, Flavio Hafner, and Megha Khosla
|
Population-Scale Network Embeddings Expose Educational Divides in Network Structure Related to Right-Wing Populist Voting
|
30 pages, 6 figures, Supplementary Materials available at https://github.com/odissei-explainable-network/netaudit; small textual changes, update Figure 6
| null | null | null |
cs.SI cs.LG stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Administrative registry data can be used to construct population-scale networks whose ties reflect shared social contexts between persons. With machine learning, such networks can be encoded into numerical representations -- embeddings -- that automatically capture individuals' position within the network. We created embeddings for all persons in the Dutch population from a population-scale network that represents five shared contexts: neighborhood, work, family, household, and school. To assess the informativeness of these embeddings, we used them to predict right-wing populist voting. Embeddings alone predicted right-wing populist voting above chance-level but performed worse than individual characteristics. Combining the best subset of embeddings with individual characteristics only slightly improved predictions. After transforming the embeddings to make their dimensions more sparse and orthogonal, we found that one embedding dimension was strongly associated with the outcome. Mapping this dimension back to the population network revealed differences in network structure related to right-wing populist voting between different school ties and achieved education levels. Our study contributes methodologically by demonstrating how population-scale network embeddings can be made interpretable, and substantively by linking structural network differences in education to right-wing populist voting.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 21:57:44 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 12:53:52 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lüken",
"Malte",
""
],
[
"Garcia-Bernardo",
"Javier",
""
],
[
"Deb",
"Sreeparna",
""
],
[
"Hafner",
"Flavio",
""
],
[
"Khosla",
"Megha",
""
]
] |
2508.21302
|
Jie Zhu
|
Jie Zhu, Chihao Shen, Ziyang Li, Jiahao Yu, Yizheng Chen, Kexin Pei
|
Locus: Agentic Predicate Synthesis for Directed Fuzzing
| null | null | null | null |
cs.CR cs.AI cs.SE
|
http://creativecommons.org/licenses/by/4.0/
|
Directed fuzzing aims to find program inputs that lead to specified target program states. It has broad applications, such as debugging system crashes, confirming reported bugs, and generating exploits for potential vulnerabilities. This task is inherently challenging because target states are often deeply nested in the program, while the search space manifested by numerous possible program inputs is prohibitively large. Existing approaches rely on branch distances or manually-specified constraints to guide the search; however, the branches alone are often insufficient to precisely characterize progress toward reaching the target states, while the manually specified constraints are often tailored for specific bug types and thus difficult to generalize to diverse target states and programs.
We present Locus, a novel framework to improve the efficiency of directed fuzzing. Our key insight is to synthesize predicates to capture fuzzing progress as semantically meaningful intermediate states, serving as milestones towards reaching the target states. When used to instrument the program under fuzzing, they can reject executions unlikely to reach the target states, while providing additional coverage guidance. To automate this task and generalize to diverse programs, Locus features an agentic framework with program analysis tools to synthesize and iteratively refine the candidate predicates, while ensuring the predicates strictly relax the target states to prevent false rejections via symbolic execution. Our evaluation shows that Locus substantially improves the efficiency of eight state-of-the-art fuzzers in discovering real-world vulnerabilities, achieving an average speedup of 41.6x. So far, Locus has found eight previously unpatched bugs, with one already acknowledged with a draft patch.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 01:47:07 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 03:33:23 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Zhu",
"Jie",
""
],
[
"Shen",
"Chihao",
""
],
[
"Li",
"Ziyang",
""
],
[
"Yu",
"Jiahao",
""
],
[
"Chen",
"Yizheng",
""
],
[
"Pei",
"Kexin",
""
]
] |
2508.21314
|
Amit Sinha
|
Amit Sinha, Matthieu Geist, Aditya Mahajan
|
Convergence of regularized agent-state-based Q-learning in POMDPs
|
Accepted in CDC 2025
| null | null | null |
cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we present a framework to understand the convergence of commonly used Q-learning reinforcement learning algorithms in practice. Two salient features of such algorithms are: (i)~the Q-table is recursively updated using an agent state (such as the state of a recurrent neural network) which is not a belief state or an information state and (ii)~policy regularization is often used to encourage exploration and stabilize the learning algorithm. We investigate the simplest form of such Q-learning algorithms which we call regularized agent-state-based Q-learning (RASQL) and show that it converges under mild technical conditions to the fixed point of an appropriately defined regularized MDP, which depends on the stationary distribution induced by the behavioral policy. We also show that a similar analysis continues to work for a variant of RASQL that learns periodic policies. We present numerical examples to illustrate that the empirical convergence behavior matches with the proposed theoretical limit.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 02:45:28 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 23:21:19 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Sinha",
"Amit",
""
],
[
"Geist",
"Matthieu",
""
],
[
"Mahajan",
"Aditya",
""
]
] |
2508.21376
|
Chi Heem Wong
|
Tony Lee, Haoqin Tu, Chi Heem Wong, Zijun Wang, Siwei Yang, Yifan Mai, Yuyin Zhou, Cihang Xie, Percy Liang
|
AHELM: A Holistic Evaluation of Audio-Language Models
| null | null | null | null |
cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Evaluations of audio-language models (ALMs) -- multimodal models that take interleaved audio and text as input and output text -- are hindered by the lack of standardized benchmarks; most benchmarks measure only one or two capabilities and omit evaluative aspects such as fairness or safety. Furthermore, comparison across models is difficult as separate evaluations test a limited number of models and use different prompting methods and inference parameters. To address these shortfalls, we introduce AHELM, a benchmark that aggregates various datasets -- including 2 new synthetic audio-text datasets called PARADE, which evaluates the ALMs on avoiding stereotypes, and CoRe-Bench, which measures reasoning over conversational audio through inferential multi-turn question answering -- to holistically measure the performance of ALMs across 10 aspects we have identified as important to the development and usage of ALMs: audio perception, knowledge, reasoning, emotion detection, bias, fairness, multilinguality, robustness, toxicity, and safety. We also standardize the prompts, inference parameters, and evaluation metrics to ensure equitable comparisons across models. We test 14 open-weight and closed-API ALMs from 3 developers and 3 additional simple baseline systems each consisting of an automatic speech recognizer and a language model. Our results show that while Gemini 2.5 Pro ranks top in 5 out of 10 aspects, it exhibits group unfairness ($p=0.01$) on ASR tasks whereas most of the other models do not. We also find that the baseline systems perform reasonably well on AHELM, with one ranking 6th overall despite having only speech-to-text capabilities. For transparency, all raw prompts, model generations, and outputs are available on our website at https://crfm.stanford.edu/helm/audio/v1.0.0. AHELM is intended to be a living benchmark and new datasets and models will be added over time.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 07:40:39 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 17:58:21 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lee",
"Tony",
""
],
[
"Tu",
"Haoqin",
""
],
[
"Wong",
"Chi Heem",
""
],
[
"Wang",
"Zijun",
""
],
[
"Yang",
"Siwei",
""
],
[
"Mai",
"Yifan",
""
],
[
"Zhou",
"Yuyin",
""
],
[
"Xie",
"Cihang",
""
],
[
"Liang",
"Percy",
""
]
] |
2508.21497
|
J.P.W. Diener
|
T. Mello, G. C. Hillhouse, and J. P. W. Diener
|
Relativistic plane wave predictions of polarization transfer observables for ($\vec{p},2\vec{p}\,$) knockout reactions from the 3s$_{1/2}$ state in $^{208}$Pb at 392 MeV
|
For v2 acknowledgements updated
| null | null | null |
nucl-th nucl-ex
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Exclusive ($\vec{p},2\vec{p}$) reactions are often described within the Relativistic Distorted Wave Impulse Approximation (RDWIA) model to include nuclear distortion effects. The Relativistic Plane Wave Impulse Approximation (RPWIA) model, on the other hand, neglects these effects on the incident and outgoing proton wave functions. Our aim is to identify kinematic regions where the RPWIA model quantitatively describes energy-sharing analyzing power (A$_y$) and qualitatively describes the unpolarized triple differential cross section ($\sigma$) data. Where no data exist, we also seek regions where RPWIA gives similar results of the complete set of polarization transfer observables (D$_{i'j}$) compared to RDWIA. The sensitivity of the RPWIA predictions to different Relativistic Mean Field (RMF) models will be assessed. For an incident 392 MeV proton beam and coplanar scattering angles, $\theta_{a'} = 32.5^\circ$ and $\theta_{b'} = -50.0^\circ$, the RPWIA predictions of $\sigma$ and A$_y$ are compared to published data for ($\vec{p},2\vec{p}$) knockout from the 3s$_{1/2}$ state in $^{208}$Pb. In the RPWIA model the proton wave functions are treated as Dirac plane waves, the $NN$ interaction is modeled using the relativistic IA1 parameterization of the Relativistic Impulse Approximation (RIA) framework, and the boundstate wave functions are extracted from the RMF models. We found that within a range of $\pm10$ MeV around the peak of $\sigma$, the RPWIA and RDWIA models quantitatively describe the A$_y$ data and qualitatively the $\sigma$. Here, the RPWIA and RDWIA predictions agree for D$_{i'j}$. While $\sigma$ is sensitive to the RMF models, D$_{i'j}$ is not. Thus concluding that RPWIA can be applied to quantitative description of D$_{i'j}$ in a limited range for proton knockout from the 3s$_{1/2}$ state in $^{208}$Pb, and qualitatively describe the shape of $\sigma$.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 10:28:16 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 13:07:52 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Mello",
"T.",
""
],
[
"Hillhouse",
"G. C.",
""
],
[
"Diener",
"J. P. W.",
""
]
] |
2508.21502
|
Marcello Porta
|
Sven Bachmann, Leonardo Goller, Marcello Porta
|
Anyons in the $\pi$-flux phase of fermionic matter coupled to a $\mathbb{Z}_2$-gauge field
|
54 pages, 21 figures
| null | null | null |
math-ph cond-mat.str-el math.MP
|
http://creativecommons.org/licenses/by/4.0/
|
We consider a system of weakly interacting spinful lattice fermions coupled to a dynamical $\mathbb{Z}_2$ gauge field. The ground state lies in the sector of a uniform $\pi$-flux per plaquette and the monopoles are massive. In the presence of a staggered mass for the fermions, this yields a fully gapped, four-dimensional ground state space on large tori. It is topologically ordered. By considering adiabatic $\pi$-flux insertion, we construct dressed monopole excitations and show that their braiding with the fermionic excitations are those of the toric code.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 10:34:55 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Bachmann",
"Sven",
""
],
[
"Goller",
"Leonardo",
""
],
[
"Porta",
"Marcello",
""
]
] |
2508.21503
|
Yunchuan Xiang
|
Yunchuan Xiang, Yingzhi Ye, Peng Feng, Xiankai Pang, Xiaofei Lan
|
A Novel Pipeline for the Identification of New Gamma-Ray Blazars from the 4FGL-Xiang Catalog Based on Multi-wavelength Flux Distributions
|
40 pages,19 figures
| null | null | null |
astro-ph.HE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Using 16 years of Fermi-LAT data and resources from the NASA/IPAC Extragalactic Database (NED), we constructed two high-precision catalogs: the updated 4FGL-Xiang-DR2 (DR2) and a supplementary version of the fifth edition of Roma-BZCAT (5BZCAT\_err), both featuring refined positional uncertainties. We performed a spatial cross-identification between the two catalogs and developed a systematic three-step analytical pipeline: (1) efficient retrieval of positional uncertainties and multi-wavelength fluxes from 5BZCAT\_err, (2) calculation of spatial association probabilities with DR2, and (3) statistical modeling of multi-band fluxes, using the Box--Cox transformation and a truncated normal distribution. Through this process, we identified 17 new blazar candidates. Fermi-LAT analysis reveals that, except for J1043.7+5323, which exhibits significant flux variability above 100 MeV, the remaining sources display stable fluxes, without apparent spatial extension or spectral curvature. Multi-wavelength flux modeling shows that 15 of the 17 sources fall within the 2$\sigma$ confidence interval of the distribution model, demonstrating strong statistical consistency with known blazar samples. As Fermi-LAT data continues to accumulate, the remaining two sources are expected to converge toward the high-confidence region, providing further support for the common-source hypothesis.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 10:37:18 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 09:57:40 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Xiang",
"Yunchuan",
""
],
[
"Ye",
"Yingzhi",
""
],
[
"Feng",
"Peng",
""
],
[
"Pang",
"Xiankai",
""
],
[
"Lan",
"Xiaofei",
""
]
] |
2508.21771
|
Gaurav Saxena
|
Srikrishna Agrawal, N. Chandnani, T. Ghosh, G. Saxena, and B. K. Agrawal, N. Paar
|
Conflation of Ensemble-Learned Nuclear Mass Models for Enhanced Precision
|
7 pages, 5 figures, 1 table in main article
| null | null | null |
nucl-th
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Ensemble learning algorithms, the gradient boosting and bagging regressors, are employed to correct the residuals of nuclear mass excess for a diverse set of six nuclear mass models. The weighted average of these corrected residuals reduces due to their partial cancellation, yielding a significant improvement in nuclear mass predictions. Our conflated model, which integrates ensemble learning and model averaging (ELMA), achieves a root mean square error of approximately 65 keV, well below the critical threshold of 100 keV, for the complete data set of Atomic Mass Evaluation (AME2020). The validity of ELMA is demonstrated through the evaluation of $Q$ values for $\alpha$ decay, showing a marked decrease in deviations from experimental data relative to predictions from individual nuclear mass models. We have also compiled a table of nuclear mass excesses and binding energies for about 6,300 nuclei, which serves as a valuable resource for various nuclear physics applications and is publicly accessible via the ELMA web interface (https://ddnp.in).
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 16:43:38 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 17:41:18 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Agrawal",
"Srikrishna",
""
],
[
"Chandnani",
"N.",
""
],
[
"Ghosh",
"T.",
""
],
[
"Saxena",
"G.",
""
],
[
"Agrawal",
"B. K.",
""
],
[
"Paar",
"N.",
""
]
] |
2508.21798
|
Sakibul Islam Sazzad
|
Rahul Dev Sharma
|
1D Cluster State Generation On Superconducting Hardware
| null | null | null | null |
quant-ph physics.app-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Measurement-based Quantum Computation(MBQC) utilize entanglement as resource for performing quantum computation. Generating cluster state using entanglement as resource is a key bottleneck for the adoption of MBQC. To generate cluster state with charge-qubit arrrays, we provide analytical derivations and numerical validations for 4-qubit cluster state. We compare our fidelities under ideal (noise-free) Hamiltonian evolution and due to effect of decoherence. We show incorporating energy relaxation ($T_1$) yields $>$90\% fidelity while pure dephasing $T_2$ show $70\%$ decays at fourth harmonics. We further show under noise $T_2$ decays to 50\% within 15 time units, versus $>$70\% under relaxation time units ($T_1$)--only. This decay quantify degradation effect of $T_2$ on preparing cluster--state preparation is more than $T_1$. We highlight the critical need for targeted error-mitigation strategies in near-term MBQC implementations.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 17:24:35 GMT"
},
{
"version": "v2",
"created": "Tue, 02 Sep 2025 20:19:22 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Sharma",
"Rahul Dev",
""
]
] |
2509.00045
|
Chong Zhang Mr.
|
Xiang Li, Chong Zhang, Hongpeng Wang, Shreyank Narayana Gowda, Yushi Li, Xiaobo Jin
|
Performance is not All You Need: Sustainability Considerations for Algorithms
|
18 pages, 6 figures. Accepted Chinese Conference on Pattern Recognition and Computer Vision 2025
| null | null | null |
cs.CV cs.PF
|
http://creativecommons.org/licenses/by/4.0/
|
This work focuses on the high carbon emissions generated by deep learning model training, specifically addressing the core challenge of balancing algorithm performance and energy consumption. It proposes an innovative two-dimensional sustainability evaluation system. Different from the traditional single performance-oriented evaluation paradigm, this study pioneered two quantitative indicators that integrate energy efficiency ratio and accuracy: the sustainable harmonic mean (FMS) integrates accumulated energy consumption and performance parameters through the harmonic mean to reveal the algorithm performance under unit energy consumption; the area under the sustainability curve (ASC) constructs a performance-power consumption curve to characterize the energy efficiency characteristics of the algorithm throughout the cycle. To verify the universality of the indicator system, the study constructed benchmarks in various multimodal tasks, including image classification, segmentation, pose estimation, and batch and online learning. Experiments demonstrate that the system can provide a quantitative basis for evaluating cross-task algorithms and promote the transition of green AI research from theory to practice. Our sustainability evaluation framework code can be found here, providing methodological support for the industry to establish algorithm energy efficiency standards.
|
[
{
"version": "v1",
"created": "Sun, 24 Aug 2025 09:53:33 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 05:15:22 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Li",
"Xiang",
""
],
[
"Zhang",
"Chong",
""
],
[
"Wang",
"Hongpeng",
""
],
[
"Gowda",
"Shreyank Narayana",
""
],
[
"Li",
"Yushi",
""
],
[
"Jin",
"Xiaobo",
""
]
] |
2509.00062
|
Justin Jung
|
Justin Jung
|
Scaffold Diffusion: Sparse Multi-Category Voxel Structure Generation with Discrete Diffusion
|
Comments: 6 pages, LaTeX; typos corrected, figure added
| null | null | null |
cs.CV cs.AI cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Generating realistic sparse multi-category 3D voxel structures is difficult due to the cubic memory scaling of voxel structures and moreover the significant class imbalance caused by sparsity. We introduce Scaffold Diffusion, a generative model designed for sparse multi-category 3D voxel structures. By treating voxels as tokens, Scaffold Diffusion uses a discrete diffusion language model to generate 3D voxel structures. We show that discrete diffusion language models can be extended beyond inherently sequential domains such as text to generate spatially coherent 3D structures. We evaluate on Minecraft house structures from the 3D-Craft dataset and demonstrate that, unlike prior baselines and an auto-regressive formulation, Scaffold Diffusion produces realistic and coherent structures even when trained on data with over 98% sparsity. We provide an interactive viewer where readers can visualize generated samples and the generation process: https://scaffold.deepexploration.org/
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 04:57:32 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 01:56:20 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Jung",
"Justin",
""
]
] |
2509.00096
|
Yao Fu
|
Yao Fu, Runchao Li, Xianxuan Long, Haotian Yu, Xiaotian Han, Yu Yin, and Pan Li
|
Pruning Weights but Not Truth: Safeguarding Truthfulness While Pruning LLMs
|
Accepted to EMNLP2025 findings (poster)
| null | null | null |
cs.LG cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Neural network pruning has emerged as a promising approach for deploying LLMs in low-resource scenarios while preserving downstream task performance. However, for the first time, we reveal that such pruning disrupts LLMs' internal activation features crucial for lie detection, where probing classifiers (typically small logistic regression models) trained on these features assess the truthfulness of LLM-generated statements. This discovery raises a crucial open question: how can we prune LLMs without sacrificing these critical lie detection capabilities? Our investigation further reveals that naively adjusting layer-wise pruning sparsity based on importance inadvertently removes crucial weights, failing to improve lie detection performance despite its reliance on the most crucial LLM layer. To address this issue, we propose Truthful Pruning aligned by Layer-wise Outliers (TPLO), which places greater emphasis on layers with more activation outliers and stronger discriminative features simultaneously. This preserves LLMs' original performance while retaining critical features of inner states needed for robust lie detection. Moreover, we introduce a prompting rule to enrich the TruthfulQA benchmark for better calibrating LLM pruning. Empirical results show that our approach improves the hallucination detection for pruned LLMs (achieving 88% accuracy at 50% sparsity) and enhances their performance on TruthfulQA.
|
[
{
"version": "v1",
"created": "Wed, 27 Aug 2025 15:48:18 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 14:58:03 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Fu",
"Yao",
""
],
[
"Li",
"Runchao",
""
],
[
"Long",
"Xianxuan",
""
],
[
"Yu",
"Haotian",
""
],
[
"Han",
"Xiaotian",
""
],
[
"Yin",
"Yu",
""
],
[
"Li",
"Pan",
""
]
] |
2509.00116
|
Akshay Kumar Jagadish
|
Akshay K. Jagadish, Mirko Thalmann, Julian Coda-Forno, Marcel Binz, and Eric Schulz
|
Meta-learning ecological priors from large language models explains human learning and decision making
| null | null | null | null |
q-bio.NC cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Human cognition is profoundly shaped by the environments in which it unfolds. Yet, it remains an open question whether learning and decision making can be explained as a principled adaptation to the statistical structure of real-world tasks. We introduce ecologically rational analysis, a computational framework that unifies the normative foundations of rational analysis with ecological grounding. Leveraging large language models to generate ecologically valid cognitive tasks at scale, and using meta-learning to derive rational models optimized for these environments, we develop a new class of learning algorithms: Ecologically Rational Meta-learned Inference (ERMI). ERMI internalizes the statistical regularities of naturalistic problem spaces and adapts flexibly to novel situations, without requiring hand-crafted heuristics or explicit parameter updates. We show that ERMI captures human behavior across 15 experiments spanning function learning, category learning, and decision making, outperforming several established cognitive models in trial-by-trial prediction. Our results suggest that much of human cognition may reflect adaptive alignment to the ecological structure of the problems we encounter in everyday life.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 16:07:40 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 03:16:09 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Jagadish",
"Akshay K.",
""
],
[
"Thalmann",
"Mirko",
""
],
[
"Coda-Forno",
"Julian",
""
],
[
"Binz",
"Marcel",
""
],
[
"Schulz",
"Eric",
""
]
] |
2509.00117
|
Jared Perlo
|
Jared Perlo, Alexander Robey, Fazl Barez, Luciano Floridi, Jakob M\"okander
|
Embodied AI: Emerging Risks and Opportunities for Policy Action
| null | null | null | null |
cs.CY cs.AI cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
The field of embodied AI (EAI) is rapidly advancing. Unlike virtual AI, EAI systems can exist in, learn from, reason about, and act in the physical world. With recent advances in AI models and hardware, EAI systems are becoming increasingly capable across wider operational domains. While EAI systems can offer many benefits, they also pose significant risks, including physical harm from malicious use, mass surveillance, as well as economic and societal disruption. These risks require urgent attention from policymakers, as existing policies governing industrial robots and autonomous vehicles are insufficient to address the full range of concerns EAI systems present. To help address this issue, this paper makes three contributions. First, we provide a taxonomy of the physical, informational, economic, and social risks EAI systems pose. Second, we analyze policies in the US, EU, and UK to assess how existing frameworks address these risks and to identify critical gaps. We conclude by offering policy recommendations for the safe and beneficial deployment of EAI systems, such as mandatory testing and certification schemes, clarified liability frameworks, and strategies to manage EAI's potentially transformative economic and societal impacts.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 17:59:07 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 17:55:11 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Perlo",
"Jared",
""
],
[
"Robey",
"Alexander",
""
],
[
"Barez",
"Fazl",
""
],
[
"Floridi",
"Luciano",
""
],
[
"Mökander",
"Jakob",
""
]
] |
2509.00142
|
Scott Scharlach
|
Scott C. Scharlach
|
Academic Miscommunication in Physics
| null | null | null | null |
physics.ed-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Academic Miscommunication (AM) is the phenomenon of a professor's expectations, beliefs, or goals in the classroom differing from those of the students. In this study, a survey was given to undergraduates in two introductory physics classes (one class intended for majors, another for non-majors) who had recently received their grade for the first exam of the semester. The survey measured whether the exam had matched the students' expectations of the exam, both in form and in content. From these responses, students were assigned an "AM score," a numerical value in which higher numbers indicate a greater deviation between the actual exam and the students' expectations. Students' exam scores were compared with their AM scores; both classes displayed an inverse linear correlation between exam scores and AM scores. The class for majors displayed R-squared = 0.291 (n=14), corresponding to a p-value of 0.031. The class for non-majors displayed an R-squared = 0.457 (n=30), corresponding to a p-value of 0.000035. The survey also inquired about several other phenomena, such as Impostor Syndrome and the growth-versus-fixed mindset paradigm; 20 graphs were plotted in total. The Bonferroni correction therefore requires a p-value of 0.0025 for statistical significance. We reject the null hypothesis concerning exam scores and feelings of AM for the physics class for non-majors. In contrast, we fail to reject the null hypothesis concerning exam scores and feelings of AM for the physics class for majors. Our findings indicate that non-physics-major students who performed poorly on the first exam of the semester had misunderstood the expectations about the style and content of the exam.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 17:59:05 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Scharlach",
"Scott C.",
""
]
] |
2509.00159
|
Matteo Boschini
|
Matteo Boschini, Davide Gerosa, Alessandro Crespi and Matteo Falcone
|
LHS in LHS: A new expansion strategy for Latin hypercube sampling in simulation design
|
8 pages, 5 figures
|
SoftwareX 31, 102294 (2025)
|
10.1016/j.softx.2025.102294
| null |
stat.ME astro-ph.HE cs.DS gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Latin Hypercube Sampling (LHS) is a prominent tool in simulation design, with a variety of applications in high-dimensional and computationally expensive problems. LHS allows for various optimization strategies, most notably to ensure space-filling properties. However, LHS is a single-stage algorithm that requires a priori knowledge of the targeted sample size. In this work, we present LHS in LHS, a new expansion algorithm for LHS that enables the addition of new samples to an existing LHS-distributed set while (approximately) preserving its properties. In summary, the algorithm identifies regions of the parameter space that are far from the initial set, draws a new LHS within those regions, and then merges it with the original samples. As a by-product, we introduce a new metric, the LHS degree, which quantifies the deviation of a given design from an LHS distribution. Our public implementation is distributed via the Python package expandLHS.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 18:00:10 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Boschini",
"Matteo",
""
],
[
"Gerosa",
"Davide",
""
],
[
"Crespi",
"Alessandro",
""
],
[
"Falcone",
"Matteo",
""
]
] |
2509.00216
|
Seongmin Kim
|
Seongmin Kim, Daihui Wu, and Yilong Han
|
Percolation transition in entangled granular networks
| null | null | null | null |
cond-mat.soft cond-mat.dis-nn cond-mat.mtrl-sci cond-mat.stat-mech
|
http://creativecommons.org/licenses/by/4.0/
|
Highly nonconvex granular particles, such as staples and metal shavings, can form solid-like cohesive structures through geometric entanglement (interlocking). The network structure formed by this entanglement, however, remains largely unexplored. Here we utilize network science to investigate the entanglement networks of C-shaped granular particles under vibration through experiments and simulations. By analyzing key network properties, we demonstrate that these networks undergo a percolation transition as the number of links increases logarithmically over time; the entangled particles form a giant cluster when the number of links exceeds a critical threshold. We propose a continuum percolation model of rings that effectively describes the observed transition. Additionally, we find that particle's opening angle significantly affects mechanical bonding and, consequently, the network structure. This work highlights the potential of network-based approaches to study entangled materials, paving the way for advancements in applications ranging from mechanical metamaterials to entangled robot swarms.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 19:58:44 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 05:41:58 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Kim",
"Seongmin",
""
],
[
"Wu",
"Daihui",
""
],
[
"Han",
"Yilong",
""
]
] |
2509.00218
|
Aleksandra Landowska
|
Aleksandra Landowska, Aislinn D Gomez Bergin, Ayodeji O. Abioye, Jayati Deshmukh, Andriana Bouadouki, Maria Wheadon, Athina Georgara, Dominic Price, Tuyen Nguyen, Shuang Ao, Lokesh Singh, Yi Long, Raffaele Miele, Joel E. Fischer, Sarvapali D. Ramchurn
|
Embodied AI in Social Spaces: Responsible and Adaptive Robots in Complex Setting -- UKAIRS 2025 (Copy)
| null | null | null | null |
cs.RO cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper introduces and overviews a multidisciplinary project aimed at developing responsible and adaptive multi-human multi-robot (MHMR) systems for complex, dynamic settings. The project integrates co-design, ethical frameworks, and multimodal sensing to create AI-driven robots that are emotionally responsive, context-aware, and aligned with the needs of diverse users. We outline the project's vision, methodology, and early outcomes, demonstrating how embodied AI can support sustainable, ethical, and human-centred futures.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 20:01:56 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Landowska",
"Aleksandra",
""
],
[
"Bergin",
"Aislinn D Gomez",
""
],
[
"Abioye",
"Ayodeji O.",
""
],
[
"Deshmukh",
"Jayati",
""
],
[
"Bouadouki",
"Andriana",
""
],
[
"Wheadon",
"Maria",
""
],
[
"Georgara",
"Athina",
""
],
[
"Price",
"Dominic",
""
],
[
"Nguyen",
"Tuyen",
""
],
[
"Ao",
"Shuang",
""
],
[
"Singh",
"Lokesh",
""
],
[
"Long",
"Yi",
""
],
[
"Miele",
"Raffaele",
""
],
[
"Fischer",
"Joel E.",
""
],
[
"Ramchurn",
"Sarvapali D.",
""
]
] |
2509.00243
|
Yves Dubief
|
Pierre-Yves Goffin, Yves Dubief, Vincent E. Terrapon
|
Follow the curvature of viscoelastic stress: Insights into the steady arrowhead structure
|
Submitted to Physical Review Fluids
| null | null | null |
cond-mat.soft physics.flu-dyn
|
http://creativecommons.org/licenses/by/4.0/
|
Focusing on simulated dilute polymer solutions, this letter investigates the interactions between flow structures and organized polymer stress sheets for the steady arrowhead coherent structure in a two-dimensional periodic channel flow. Formulating the problem in a frame of reference moving with the arrowhead velocity, streamlines, which are also pathlines in this frame, enables the identification of two distinct topological regions linked to two stagnation points. The streamlines help connecting the spatial distribution of polymer stress within the sheets and the dynamics of polymers transported by the flow. Using stresslines, lines parallel to the eigenvectors of polymer stress, a novel formulation of the viscoelastic stress term in the momentum transport equation proposes a more intuitive interpretation of the relation between the curvature of the stresslines, and the variation of stress along these lines, with the local flow topology. An approximation of this formulation is shown to explain the pressure jump observed in the arrowhead structure as a function of the local curvature of the polymer stress sheet.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 21:20:32 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Goffin",
"Pierre-Yves",
""
],
[
"Dubief",
"Yves",
""
],
[
"Terrapon",
"Vincent E.",
""
]
] |
2509.00364
|
Shu Zhang
|
Shu Zhang, J. R. Fuentes, Andrew Cumming
|
3D Simulations of Convective Entrainment in Gas Giants: Rotation and Decreasing Luminosity as Barriers to Mixing
|
Accepted in ApJ Letters
| null | null | null |
astro-ph.EP
|
http://creativecommons.org/licenses/by/4.0/
|
Observations from Juno and Cassini suggest that Jupiter and Saturn may possess fuzzy cores -- central regions where the abundance of heavy elements varies smoothly with depth. Such gradients pose a longstanding puzzle for models of planetary evolution and formation, which predict that vigorous convection would homogenize the interior of gas giants within the first $\sim 10^6$--$10^8~\mathrm{years}$ of cooling. Previous 3D simulations and analytic predictions for the propagation of a convection zone into a stable region have demonstrated that the rapid rotation of gas giants can significantly slow convective mixing, but not enough to stop it. Another piece of the puzzle is luminosity. Gas giants cool as they age, and with that comes a declining heat flux over time. Recent ideas suggest that when this declining luminosity is combined with rotational effects, convection may stall. We explore this possibility using 3D hydrodynamic simulations that include both rotation and a surface cooling flux that decreases as $1/t$. Our results demonstrate that, even without rotation, a declining luminosity can suppress mixing sufficiently to preserve an initial compositional gradient in the deep interior of gas giants. If confirmed by more realistic simulations, this may help to explain the long-term survival of fuzzy cores.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 05:22:50 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 15:39:07 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Zhang",
"Shu",
""
],
[
"Fuentes",
"J. R.",
""
],
[
"Cumming",
"Andrew",
""
]
] |
2509.00367
|
Numan Saeed
|
Numan Saeed, Salma Hassan, Shahad Hardan, Ahmed Aly, Darya Taratynova, Umair Nawaz, Ufaq Khan, Muhammad Ridzuan, Vincent Andrearczyk, Adrien Depeursinge, Mathieu Hatt, Thomas Eugene, Rapha\"el Metz, M\'elanie Dore, Gregory Delpon, Vijay Ram Kumar Papineni, Kareem Wahid, Cem Dede, Alaa Mohamed Shawky Ali, Carlos Sjogreen, Mohamed Naser, Clifton D. Fuller, Valentin Oreiller, Mario Jreige, John O. Prior, Catherine Cheze Le Rest, Olena Tankyevych, Pierre Decazes, Su Ruan, Stephanie Tanadini-Lang, Martin Valli\`eres, Hesham Elhalawani, Ronan Abgral, Romain Floch, Kevin Kerleguer, Ulrike Schick, Maelle Mauguen, Arman Rahmim, Mohammad Yaqub
|
A Multimodal and Multi-centric Head and Neck Cancer Dataset for Tumor Segmentation and Outcome Prediction
|
10 pages, 5 figures. Numan Saeed is the corresponding author. Numan Saeed, Salma Hassan and Shahad Hardan contributed equally to this work. Project page: https://hecktor25.grand-challenge.org/
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
We describe a publicly available multimodal dataset of annotated Positron Emission Tomography/Computed Tomography (PET/CT) studies for head and neck cancer research. The dataset includes 1123 FDG-PET/CT studies from patients with histologically confirmed head and neck cancer, acquired from 10 international medical centers. All examinations consisted of co-registered PET/CT scans with varying acquisition protocols, reflecting real-world clinical diversity across institutions. Primary gross tumor volumes (GTVp) and involved lymph nodes (GTVn) were manually segmented by experienced radiation oncologists and radiologists following standardized guidelines and quality control measures. We provide anonymized NifTi files of all studies, along with expert-annotated segmentation masks, radiotherapy dose distribution for a subset of patients, and comprehensive clinical metadata. This metadata includes TNM staging, HPV status, demographics (age and gender), long-term follow-up outcomes, survival times, censoring indicators, and treatment information. We demonstrate how this dataset can be used for three key clinical tasks: automated tumor segmentation, recurrence-free survival prediction, and HPV status classification, providing benchmark results using state-of-the-art deep learning models, including UNet, SegResNet, and multimodal prognostic frameworks.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 05:38:48 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 04:28:00 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Saeed",
"Numan",
""
],
[
"Hassan",
"Salma",
""
],
[
"Hardan",
"Shahad",
""
],
[
"Aly",
"Ahmed",
""
],
[
"Taratynova",
"Darya",
""
],
[
"Nawaz",
"Umair",
""
],
[
"Khan",
"Ufaq",
""
],
[
"Ridzuan",
"Muhammad",
""
],
[
"Andrearczyk",
"Vincent",
""
],
[
"Depeursinge",
"Adrien",
""
],
[
"Hatt",
"Mathieu",
""
],
[
"Eugene",
"Thomas",
""
],
[
"Metz",
"Raphaël",
""
],
[
"Dore",
"Mélanie",
""
],
[
"Delpon",
"Gregory",
""
],
[
"Papineni",
"Vijay Ram Kumar",
""
],
[
"Wahid",
"Kareem",
""
],
[
"Dede",
"Cem",
""
],
[
"Ali",
"Alaa Mohamed Shawky",
""
],
[
"Sjogreen",
"Carlos",
""
],
[
"Naser",
"Mohamed",
""
],
[
"Fuller",
"Clifton D.",
""
],
[
"Oreiller",
"Valentin",
""
],
[
"Jreige",
"Mario",
""
],
[
"Prior",
"John O.",
""
],
[
"Rest",
"Catherine Cheze Le",
""
],
[
"Tankyevych",
"Olena",
""
],
[
"Decazes",
"Pierre",
""
],
[
"Ruan",
"Su",
""
],
[
"Tanadini-Lang",
"Stephanie",
""
],
[
"Vallières",
"Martin",
""
],
[
"Elhalawani",
"Hesham",
""
],
[
"Abgral",
"Ronan",
""
],
[
"Floch",
"Romain",
""
],
[
"Kerleguer",
"Kevin",
""
],
[
"Schick",
"Ulrike",
""
],
[
"Mauguen",
"Maelle",
""
],
[
"Rahmim",
"Arman",
""
],
[
"Yaqub",
"Mohammad",
""
]
] |
2509.00379
|
Jialiang Kang
|
Jialiang Kang, Jiawen Wang, Dingsheng Luo
|
Domain Adaptation-Based Crossmodal Knowledge Distillation for 3D Semantic Segmentation
|
ICRA 2025
| null | null | null |
cs.CV cs.RO
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Semantic segmentation of 3D LiDAR data plays a pivotal role in autonomous driving. Traditional approaches rely on extensive annotated data for point cloud analysis, incurring high costs and time investments. In contrast, realworld image datasets offer abundant availability and substantial scale. To mitigate the burden of annotating 3D LiDAR point clouds, we propose two crossmodal knowledge distillation methods: Unsupervised Domain Adaptation Knowledge Distillation (UDAKD) and Feature and Semantic-based Knowledge Distillation (FSKD). Leveraging readily available spatio-temporally synchronized data from cameras and LiDARs in autonomous driving scenarios, we directly apply a pretrained 2D image model to unlabeled 2D data. Through crossmodal knowledge distillation with known 2D-3D correspondence, we actively align the output of the 3D network with the corresponding points of the 2D network, thereby obviating the necessity for 3D annotations. Our focus is on preserving modality-general information while filtering out modality-specific details during crossmodal distillation. To achieve this, we deploy self-calibrated convolution on 3D point clouds as the foundation of our domain adaptation module. Rigorous experimentation validates the effectiveness of our proposed methods, consistently surpassing the performance of state-of-the-art approaches in the field.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 06:34:39 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Kang",
"Jialiang",
""
],
[
"Wang",
"Jiawen",
""
],
[
"Luo",
"Dingsheng",
""
]
] |
2509.00399
|
Ritwik Mukherjee
|
Ritwik Mukherjee, Abhishek Dhar, Manas Kulkarni, Samriddhi Sankar Ray
|
Extreme dynamics and relaxation of quantum gases: A hydrodynamic approach
|
13 pages, 8 figures
| null | null | null |
cond-mat.quant-gas cond-mat.stat-mech physics.flu-dyn
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The evolution of quantum gases, released from traps, are studied through hydrodynamics, both analytically and numerically, in one and two dimensions. In particular, we demonstrate the existence of long time self-similar solutions of the Euler equations, for the density and velocity fields, and derive the scaling exponents as well as the scaling functions. We find that the expanding gas develops a shock front and the size of the cloud grows in time as a powerlaw. We relate the associated exponent to that appearing in the corresponding equation of state of the quantum gas. Furthermore, we study the relaxation dynamics of a trapped quantum gas and show that the resulting steady state is in excellent agreement with that derived analytically. Our hydrodynamic approach is versatile and can be used to unravel several other far-from-equilibrium collective phenomenon of extreme nature, relevant to the growing experimental interests in quantum gases.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 07:49:23 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Mukherjee",
"Ritwik",
""
],
[
"Dhar",
"Abhishek",
""
],
[
"Kulkarni",
"Manas",
""
],
[
"Ray",
"Samriddhi Sankar",
""
]
] |
2509.00409
|
Shanmugasundaram Sundar
|
C.H. Namitha, S. Sundar, Shankar Veerabathiran
|
Doubly Commuting Semigroups of Isometries
| null | null | null | null |
math.OA math.FA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we discuss the structure of doubly commuting semigroups of isometries. We record a new proof of Cooper's theorem in the Hilbert module setting. We discuss the Fell topology on the set of equivalence classes of irreducible, doubly commuting isometric representations of $\mathbb{R}_{+}^{d}$. We show that if $d$ is finite, the topology is $T_0$. We indicate the pathologies that occur when $d=\infty$. In particular, we show that Wold decomposition fails for isometric representations of $\mathbb{R}_{+}^{\infty}$ and prove that the Fell topology on the set of equivalence classes of irreducible, doubly commuting isometric representations of $\mathbb{R}_{+}^{\infty}$ is not $T_0$
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 08:35:16 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 06:23:25 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Namitha",
"C. H.",
""
],
[
"Sundar",
"S.",
""
],
[
"Veerabathiran",
"Shankar",
""
]
] |
2509.00438
|
Feng-Yu Lu
|
Feng-Yu Lu, Jia-Xuan Li, Ze-Hao Wang, Shuang Wang, Zhen-Qiang Yin, Alvaro Navarrete, Marcos Curty, Wei Chen, De-Yong He, Guang-Can Guo, Zheng-Fu Han
|
Breaking the system-frequency limitation of quantum key distribution
|
34 pages, 11 figures
| null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Enhancing the system performance has been the primary mission in the domain of quantum key distributions in recent years. Higher performance necessitates a higher repetition frequency, which is, however, strictly limited by the bandwidth. Excessive increase in the repetition frequency is not justifiable since it leads to misaligned and correlated modulation, resulting in compromised security and significant errors. Meanwhile, replacing devices with higher bandwidth means more technique challenges and more expense, which is detrimental to practical implementations. In this work, we propose a comprehensive countermeasure to overcome the bandwidth limitation. We present a new protocol addressing the aforementioned loopholes to ensure security and achieve superior performance compared to previous approaches. Additionally, we address the challenges associated with characterizing and mitigating the correlations by developing two techniques. An experiment setup is completed in this work, whose result indicates that our technique achieves the lowest correlated deviation in all similar works and our proposal breaks the bandwidth-limited secret key rate, thus releasing people from the trade-off between performance and cost, paving the way for practical applications.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 09:56:30 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 06:48:59 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lu",
"Feng-Yu",
""
],
[
"Li",
"Jia-Xuan",
""
],
[
"Wang",
"Ze-Hao",
""
],
[
"Wang",
"Shuang",
""
],
[
"Yin",
"Zhen-Qiang",
""
],
[
"Navarrete",
"Alvaro",
""
],
[
"Curty",
"Marcos",
""
],
[
"Chen",
"Wei",
""
],
[
"He",
"De-Yong",
""
],
[
"Guo",
"Guang-Can",
""
],
[
"Han",
"Zheng-Fu",
""
]
] |
2509.00469
|
Houyi Sun
|
Houyi Sun and Ya-Ping Li and Zhen Pan and Huan Yang
|
Probing Formation Channels of Extreme Mass-Ratio Inspirals
|
37 pages, 30 figures, 2 tables
| null | null | null |
gr-qc astro-ph.HE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The population study of stellar-mass black hole (sBH) binaries with ground-based gravitational wave detection has achieved tremendous success in recent years. Future observation of extreme mass-ratio inspirals will similarly require proper population analysis that identify the formation channels, measuring the branch ratio(s) and characterizing major properties within each major channel. In this work, we propose that the measurement of eccentricity, inclination, and component mass provides critical information to distinguish different formation channels and probe detailed formation mechanisms. Focusing on the dry and wet extreme mass-ratio inspirals, we establish the theoretical expectation of these observables in each formation channel. We also discuss how their distributions can be used to probe lifetime and turbulence level of active galactic nuclei disks, accretion patterns of supermassive black holes and population properties of sBHs within nuclear star clusters.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 12:07:02 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Sun",
"Houyi",
""
],
[
"Li",
"Ya-Ping",
""
],
[
"Pan",
"Zhen",
""
],
[
"Yang",
"Huan",
""
]
] |
2509.00578
|
ABDELLAH Zakaria Sellam Mr
|
Abdellah Zakaria Sellam, Ilyes Benaissa, Salah Eddine Bekhouche, Abdenour Hadid, Vito Ren\'o, Cosimo Distante
|
C-DiffDet+: Fusing Global Scene Context with Generative Denoising for High-Fidelity Object Detection
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Fine-grained object detection in challenging visual domains, such as vehicle damage assessment, presents a formidable challenge even for human experts to resolve reliably. While DiffusionDet has advanced the state-of-the-art through conditional denoising diffusion, its performance remains limited by local feature conditioning in context-dependent scenarios. We address this fundamental limitation by introducing Context-Aware Fusion (CAF), which leverages cross-attention mechanisms to integrate global scene context with local proposal features directly. The global context is generated using a separate dedicated encoder that captures comprehensive environmental information, enabling each object proposal to attend to scene-level understanding. Our framework significantly enhances the generative detection paradigm by enabling each object proposal to attend to comprehensive environmental information. Experimental results demonstrate an improvement over state-of-the-art models on the CarDD benchmark, establishing new performance benchmarks for context-aware object detection in fine-grained domains
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 18:06:06 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 15:41:02 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Sellam",
"Abdellah Zakaria",
""
],
[
"Benaissa",
"Ilyes",
""
],
[
"Bekhouche",
"Salah Eddine",
""
],
[
"Hadid",
"Abdenour",
""
],
[
"Renó",
"Vito",
""
],
[
"Distante",
"Cosimo",
""
]
] |
2509.00587
|
Vaibhav Mehta
|
Vaibhav Mehta, Justin Hsu
|
A Hoare Logic for Symmetry Properties
|
Accepted to OOPSLA '25
| null | null | null |
cs.PL
|
http://creativecommons.org/licenses/by/4.0/
|
Many natural program correctness properties can be stated in terms of symmetries, but existing formal methods have little support for reasoning about such properties. We consider how to formally verify a broad class of symmetry properties expressed in terms of group actions. To specify these properties, we design a syntax for group actions, supporting standard constructions and a natural notion of entailment. Then, we develop a Hoare-style logic for verifying symmetry properties of imperative programs, where group actions take the place of the typical pre- and post-condition assertions. Finally, we develop a prototype tool SymVerif, and use it to verify symmetry properties on a series of handcrafted benchmarks. Our tool uncovered an error in a model of a dynamical system described by \citet{McLachlan_Quispel_2002}.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 18:46:28 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 02:59:29 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Mehta",
"Vaibhav",
""
],
[
"Hsu",
"Justin",
""
]
] |
2509.00591
|
Lang Xiong
|
Lang Xiong, Nishant Bhargava, Wesley Chang, Jianhang Hong, Haihao Liu, and Kevin Zhu
|
Probe-Rewrite-Evaluate: A Workflow for Reliable Benchmarks and Quantifying Evaluation Awareness
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Large Language Models (LLMs) often exhibit significant behavioral shifts when they perceive a change from a real-world deployment context to a controlled evaluation setting, a phenomenon known as "evaluation awareness." This discrepancy poses a critical challenge for AI alignment, as benchmark performance may not accurately reflect a model's true safety and honesty. In this work, we systematically quantify these behavioral changes by manipulating the perceived context of prompts. We introduce a methodology that uses a linear probe to score prompts on a continuous scale from "test-like" to "deploy-like" and leverage an LLM rewriting strategy to shift these prompts towards a more natural, deployment-style context while preserving the original task. Using this method, we achieved a 30% increase in the average probe score across a strategic role-playing dataset after rewriting. Evaluating a suite of state-of-the-art models on these original and rewritten prompts, we find that rewritten "deploy-like" prompts induce a significant and consistent shift in behavior. Across all models, we observed an average increase in honest responses of 5.26% and a corresponding average decrease in deceptive responses of 12.40%. Furthermore, refusal rates increased by an average of 6.38%, indicating heightened safety compliance. Our findings demonstrate that evaluation awareness is a quantifiable and manipulable factor that directly influences LLM behavior, revealing that models are more prone to unsafe or deceptive outputs in perceived test environments. This underscores the urgent need for more realistic evaluation frameworks to accurately gauge true model alignment before deployment.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 19:03:14 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 06:00:33 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Xiong",
"Lang",
""
],
[
"Bhargava",
"Nishant",
""
],
[
"Chang",
"Wesley",
""
],
[
"Hong",
"Jianhang",
""
],
[
"Liu",
"Haihao",
""
],
[
"Zhu",
"Kevin",
""
]
] |
2509.00609
|
Runcao Lyu
|
Runcao Lyu
|
$C^{\infty}$ Regularity for the free boundary of one-phase Fractional Laplacian problem
| null | null | null | null |
math.AP
|
http://creativecommons.org/licenses/by/4.0/
|
We consider a one-phase free boundary problem involving fractional Laplacian $(-\Delta)^s$, $0<s<1$. D. De Silva, O. Savin, and Y. Sire proved that the flat boundaries are $C^{1,\alpha}$. We raise the regularity to $C^{\infty}$, extending the result known for $(-\Delta)^{1/2}$ by D. De Silva and O. Savin.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 21:17:29 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:55:20 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Lyu",
"Runcao",
""
]
] |
2509.00638
|
Ce Xu
|
Hongyuan Rui and Ce Xu
|
Contour Integrations and Parity Results of Cyclotomic Euler Sums and Multiple Polylogarithm Function
|
26 pages
| null | null | null |
math.NT
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we define extended trigonometric functions via series and employ the method of contour integration to investigate the parity of certain cyclotomic Euler sums and multiple polylogarithm function. We can provide the statement of parity results for cyclotomic Euler sums of arbitrary order, explicit formulas for the parity of cyclotomic linear and quadratic Euler sums, as well as some formulas for the parity of cyclotomic cubic Euler sums and multiple polylogarithms. As a direct corollary, we derive known formulas concerning the parity of classical Euler sums and alternating Euler sums.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 23:56:16 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 01:15:20 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Rui",
"Hongyuan",
""
],
[
"Xu",
"Ce",
""
]
] |
2509.00665
|
Weilong Yan
|
Weilong Yan, Xin Zhang, Robby T. Tan
|
ER-LoRA: Effective-Rank Guided Adaptation for Weather-Generalized Depth Estimation
| null | null | null | null |
cs.CV cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
Monocular depth estimation under adverse weather conditions (e.g.\ rain, fog, snow, and nighttime) remains highly challenging due to the lack of reliable ground truth and the difficulty of learning from unlabeled real-world data. Existing methods often rely on synthetic adverse data with pseudo-labels, which suffer from domain gaps, or employ self-supervised learning, which violates photometric assumptions in adverse scenarios. In this work, we propose to achieve weather-generalized depth estimation by Parameter-Efficient Fine-Tuning (PEFT) of Vision Foundation Models (VFMs), using only a small amount of high-visibility (normal) data. While PEFT has shown strong performance in semantic tasks such as segmentation, it remains underexplored for geometry -- centric tasks like depth estimation -- especially in terms of balancing effective adaptation with the preservation of pretrained knowledge. To this end, we introduce the Selecting-Tuning-Maintaining (STM) strategy, which structurally decomposes the pretrained weights of VFMs based on two kinds of effective ranks (entropy-rank and stable-rank). In the tuning phase, we adaptively select the proper rank number as well as the task-aware singular directions for initialization, based on the entropy-rank and full-tuned weight; while in the maintaining stage, we enforce a principal direction regularization based on the stable-rank. This design guarantees flexible task adaptation while preserving the strong generalization capability of the pretrained VFM. Extensive experiments on four real-world benchmarks across diverse weather conditions demonstrate that STM not only outperforms existing PEFT methods and full fine-tuning but also surpasses methods trained with adverse synthetic data, and even the depth foundation model
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 02:24:00 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Yan",
"Weilong",
""
],
[
"Zhang",
"Xin",
""
],
[
"Tan",
"Robby T.",
""
]
] |
2509.00694
|
Xiaoping Zhai
|
Tao Liang, Jiahong Wu, and Xiaoping Zhai
|
Improved stability threshold for 2D Navier-Stokes Couette flow in an infinite channel
|
Any comments are welcome
| null | null | null |
math.AP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We study the nonlinear stability of the two-dimensional Navier-Stokes equations around the Couette shear flow in the channel domain $\mathbb{R}\times[-1,1]$ subject to Navier slip boundary conditions. We establish a quantitative stability threshold for perturbations of the initial vorticity $\omega_{in}$, showing that stability holds for perturbations of order $\nu^{1/2}$ measured in an anisotropic Sobolev space. This sharpens the recent work of Arbon and Bedrossian [Comm. Math. Phys., 406 (2025), Paper No. 129] who proved stability under the threshold $\nu^{1/2}(1+\ln(1/\nu))^{-1/2}$. Our result removes the logarithmic loss and identifies the natural scaling $\nu^{1/2}$ as the critical size of perturbations for nonlinear stability in this setting.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 04:19:13 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 03:56:57 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Liang",
"Tao",
""
],
[
"Wu",
"Jiahong",
""
],
[
"Zhai",
"Xiaoping",
""
]
] |
2509.00761
|
Boqin Yuan
|
Ziqi Wang, Boqin Yuan
|
L-MARS: Legal Multi-Agent Workflow with Orchestrated Reasoning and Agentic Search
| null | null | null | null |
cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
We present L-MARS (Legal Multi-Agent Workflow with Orchestrated Reasoning and Agentic Search), a system that reduces hallucination and uncertainty in legal question answering through coordinated multi-agent reasoning and retrieval. Unlike single-pass retrieval-augmented generation (RAG), L-MARS decomposes queries into subproblems, issues targeted searches across heterogeneous sources (Serper web, local RAG, CourtListener case law), and employs a Judge Agent to verify sufficiency, jurisdiction, and temporal validity before answer synthesis. This iterative reasoning-search-verification loop maintains coherence, filters noisy evidence, and grounds answers in authoritative law. We evaluated L-MARS on LegalSearchQA, a new benchmark of 200 up-to-date multiple choice legal questions in 2025. Results show that L-MARS substantially improves factual accuracy, reduces uncertainty, and achieves higher preference scores from both human experts and LLM-based judges. Our work demonstrates that multi-agent reasoning with agentic search offers a scalable and reproducible blueprint for deploying LLMs in high-stakes domains requiring precise legal retrieval and deliberation.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 09:23:26 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 00:57:14 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Wang",
"Ziqi",
""
],
[
"Yuan",
"Boqin",
""
]
] |
2509.00769
|
Dandan Fan
|
Dandan Fan, Huiqiu Lin, Yinfen Zhu
|
A note on the spectral radius and $[a,b]$-factor of graphs
| null | null | null | null |
math.SP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The investigation of eigenvalue conditions for the existence of an $[a,b]$-factor originates in the work of Brouwer and Haemers (2005) on perfect matchings. In the decades since, spectral extremal problems related to $[a,b]$-factors have attracted considerable attention. In this paper, we establish a spectral radius condition that ensures the existence of an $[a,b]$-factor in a graph $G$ with minimum degree $\delta(G) \geq a$, where $b > a \geq 1$. This result resolves a problem posed by Hao and Li [Electron. J. Combin. (2024)].
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 09:47:17 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 09:31:05 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Fan",
"Dandan",
""
],
[
"Lin",
"Huiqiu",
""
],
[
"Zhu",
"Yinfen",
""
]
] |
2509.00798
|
ChangIn Choi
|
Changin Choi, Wonseok Lee, Jungmin Ko, Wonjong Rhee
|
Multimodal Iterative RAG for Knowledge Visual Question Answering
| null | null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
While Multimodal Large Language Models (MLLMs) have significantly advanced multimodal understanding, their performance remains limited on knowledge-intensive visual questions that require external knowledge beyond the image. Retrieval-Augmented Generation (RAG) has become a promising solution for providing models with external knowledge, its conventional single-pass framework often fails to gather sufficient knowledge. To overcome this limitation, we propose MI-RAG, a Multimodal Iterative RAG framework that leverages reasoning to enhance retrieval and update reasoning over newly retrieved knowledge across modalities. At each iteration, MI-RAG leverages an accumulated reasoning record to dynamically formulate a multi-query. These queries then drive a joint search across heterogeneous knowledge bases containing both visually-grounded and textual knowledge. The newly acquired knowledge is synthesized into the reasoning record, progressively refining understanding across iterations. Experiments on challenging benchmarks, including Encyclopedic VQA, InfoSeek, and OK-VQA, show that MI-RAG significantly improves both retrieval recall and answer accuracy, establishing a scalable approach for compositional reasoning in knowledge-intensive VQA.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 11:14:54 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 14:36:06 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Choi",
"Changin",
""
],
[
"Lee",
"Wonseok",
""
],
[
"Ko",
"Jungmin",
""
],
[
"Rhee",
"Wonjong",
""
]
] |
2509.00831
|
Zhijing Wu
|
Zhijing Wu, Longguang Wang
|
UPGS: Unified Pose-aware Gaussian Splatting for Dynamic Scene Deblurring
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Reconstructing dynamic 3D scenes from monocular video has broad applications in AR/VR, robotics, and autonomous navigation, but often fails due to severe motion blur caused by camera and object motion. Existing methods commonly follow a two-step pipeline, where camera poses are first estimated and then 3D Gaussians are optimized. Since blurring artifacts usually undermine pose estimation, pose errors could be accumulated to produce inferior reconstruction results. To address this issue, we introduce a unified optimization framework by incorporating camera poses as learnable parameters complementary to 3DGS attributes for end-to-end optimization. Specifically, we recast camera and object motion as per-primitive SE(3) affine transformations on 3D Gaussians and formulate a unified optimization objective. For stable optimization, we introduce a three-stage training schedule that optimizes camera poses and Gaussians alternatively. Particularly, 3D Gaussians are first trained with poses being fixed, and then poses are optimized with 3D Gaussians being untouched. Finally, all learnable parameters are optimized together. Extensive experiments on the Stereo Blur dataset and challenging real-world sequences demonstrate that our method achieves significant gains in reconstruction quality and pose estimation accuracy over prior dynamic deblurring methods.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 13:01:03 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 16:49:23 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Wu",
"Zhijing",
""
],
[
"Wang",
"Longguang",
""
]
] |
2509.00856
|
Alexander Komech
|
A.I. Komech and E.A. Kopylova
|
On dissipation operators of Quantum Optics
|
5 pages, 0 figures
| null | null | null |
quant-ph math-ph math.MP
|
http://creativecommons.org/licenses/by/4.0/
|
We consider dissipation operators used in Quantum Optics for the description of quantum spontaneous emission in the context of damped driven Jaynes-Cummings equations. The equations describe quantised one-mode Maxwell field coupled to a two-level molecule. Our main result is the symmetry and nonpositivity of basic dissipation operator of Quantum Optics.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 14:02:41 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 07:11:43 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Komech",
"A. I.",
""
],
[
"Kopylova",
"E. A.",
""
]
] |
2509.00868
|
Yong Su
|
Yong Su, Yiyi Chen, Shenghong Yi, Hui Feng, Yuedong Xu, Wang Xiang, Bo Hu
|
A Modular and Scalable Simulator for Connected-UAVs Communication in 5G Networks
|
a short version is accepted by MSWiM 2025
| null | null | null |
cs.NI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Cellular-connected UAV systems have enabled a wide range of low-altitude aerial services. However, these systems still face many challenges, such as frequent handovers and the inefficiency of traditional transport protocols. To better study these issues, we develop a modular and scalable simulation platform specifically designed for UAVs communication leveraging the research ecology in wireless communication of MATLAB. The platform supports flexible 5G NR node deployment, customizable UAVs mobility models, and multi-network-interface extensions. It also supports multiple transport protocols including TCP, UDP, QUIC, etc., allowing to investigate how different transport protocols affect UAVs communication performance. In addition, the platform includes a handover management module, enabling the evaluation of both traditional and learning-based handover strategies. Our platform can serve as a testbed for the development and evaluation of advanced transmission strategies in cellular-connected UAV systems.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 14:29:50 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 03:56:12 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Su",
"Yong",
""
],
[
"Chen",
"Yiyi",
""
],
[
"Yi",
"Shenghong",
""
],
[
"Feng",
"Hui",
""
],
[
"Xu",
"Yuedong",
""
],
[
"Xiang",
"Wang",
""
],
[
"Hu",
"Bo",
""
]
] |
2509.00873
|
Vimal Kumar
|
Vimal Kumar, Susmita Das, and Mala Das
|
The Dark Matter Search at Jaduguda Underground Science Laboratory
|
Contribution to : PPC 2024
| null | null | null |
hep-ex
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
The first run for the dark matter direct search experiment at Jaduguda Underground Science Laboratory is presented in this article. The experiment named InDEx; the Indian Dark matter search Experiment has been initiated with superheated emulsion detector consisting of the droplets of tetrafluoroethane ($C_2H_2F_4$). The detector ran for an effective period of 48.6 days at a threshold of 5.87 keV with an exposure of 2.47 kg-days. It is observed that the minimum sensitivity for the SI cross-section of $[7.939\pm(0.375)_{statistical}(^{+1.386}_{-0.909})_{systematic}]\times 10^{-39}$ $cm^2$ for fluorine appears at WIMP mass of 30.67 $GeV/c^2$. The InDEx with larger exposures is under development and yet to come in near future.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 14:38:52 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 11:25:47 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Kumar",
"Vimal",
""
],
[
"Das",
"Susmita",
""
],
[
"Das",
"Mala",
""
]
] |
2509.00887
|
Takuya Miyamoto
|
Takuya Miyamoto
|
Hasse-Schmidt derivations and locally trivial deformations in positive characteristic
|
46 pages
| null | null | null |
math.AG math.AC
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we first study the lifting problem of Hasse-Schmidt derivations and then apply the results to locally trivial deformations of algebraic schemes, both in positive characteristic setting. In particular, we show that any locally trivial deformation of an affine scheme is trivial even in positive characteristic.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 14:56:47 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Miyamoto",
"Takuya",
""
]
] |
2509.00891
|
Zonghai Yao
|
Zonghai Yao, Talha Chafekar, Junda Wang, Shuo Han, Feiyun Ouyang, Junhui Qian, Lingxi Li, Hong Yu
|
ChatCLIDS: Simulating Persuasive AI Dialogues to Promote Closed-Loop Insulin Adoption in Type 1 Diabetes Care
|
Equal contribution for the first two authors
| null | null | null |
cs.AI cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Real-world adoption of closed-loop insulin delivery systems (CLIDS) in type 1 diabetes remains low, driven not by technical failure, but by diverse behavioral, psychosocial, and social barriers. We introduce ChatCLIDS, the first benchmark to rigorously evaluate LLM-driven persuasive dialogue for health behavior change. Our framework features a library of expert-validated virtual patients, each with clinically grounded, heterogeneous profiles and realistic adoption barriers, and simulates multi-turn interactions with nurse agents equipped with a diverse set of evidence-based persuasive strategies. ChatCLIDS uniquely supports longitudinal counseling and adversarial social influence scenarios, enabling robust, multi-dimensional evaluation. Our findings reveal that while larger and more reflective LLMs adapt strategies over time, all models struggle to overcome resistance, especially under realistic social pressure. These results highlight critical limitations of current LLMs for behavior change, and offer a high-fidelity, scalable testbed for advancing trustworthy persuasive AI in healthcare and beyond.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 15:08:41 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 04:55:57 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Yao",
"Zonghai",
""
],
[
"Chafekar",
"Talha",
""
],
[
"Wang",
"Junda",
""
],
[
"Han",
"Shuo",
""
],
[
"Ouyang",
"Feiyun",
""
],
[
"Qian",
"Junhui",
""
],
[
"Li",
"Lingxi",
""
],
[
"Yu",
"Hong",
""
]
] |
2509.00905
|
MAOyuan Shao
|
Yutong Gao, Maoyuan Shao, Xinyang Huang, Chuang Zhu, Lijuan Sun, Yu Weng, Xuan Liu, Guoshun Nan
|
Spotlighter: Revisiting Prompt Tuning from a Representative Mining View
|
Accepted as EMNLP 2025 Findings
|
EMNLP2025
| null | null |
cs.CV cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
CLIP's success has demonstrated that prompt tuning can achieve robust cross-modal semantic alignment for tasks ranging from open-domain recognition to fine-grained classification. However, redundant or weakly relevant feature components introduce noise and incur unnecessary computational costs. In this work, we propose Spotlighter, a lightweight token-selection framework that simultaneously enhances accuracy and efficiency in prompt tuning. Spotlighter evaluates each visual token's activation from both sample-wise and semantic-wise perspectives and retains only the top-scoring tokens for downstream prediction. A class-specific semantic memory bank of learned prototypes refines this selection, ensuring semantic representativeness and compensating for discarded features. To further prioritize informative signals, we introduce a two-level ranking mechanism that dynamically weights token--prototype interactions. Across 11 few-shot benchmarks, Spotlighter outperforms CLIP by up to 11.19\% in harmonic mean accuracy and achieves up to 0.8K additional FPS, with only 21 extra parameters. These results establish Spotlighter as an effective and scalable baseline for prompt tuning. Code for our method will be available at https://github.com/greatest-gourmet/Spotlighter.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 15:37:48 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 01:19:20 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Gao",
"Yutong",
""
],
[
"Shao",
"Maoyuan",
""
],
[
"Huang",
"Xinyang",
""
],
[
"Zhu",
"Chuang",
""
],
[
"Sun",
"Lijuan",
""
],
[
"Weng",
"Yu",
""
],
[
"Liu",
"Xuan",
""
],
[
"Nan",
"Guoshun",
""
]
] |
2509.00911
|
Joongho Jo
|
Joongho Jo, Jongsun Park
|
GS-TG: 3D Gaussian Splatting Accelerator with Tile Grouping for Reducing Redundant Sorting while Preserving Rasterization Efficiency
|
DAC 2025
| null | null | null |
cs.AR cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
3D Gaussian Splatting (3D-GS) has emerged as a promising alternative to neural radiance fields (NeRF) as it offers high speed as well as high image quality in novel view synthesis. Despite these advancements, 3D-GS still struggles to meet the frames per second (FPS) demands of real-time applications. In this paper, we introduce GS-TG, a tile-grouping-based accelerator that enhances 3D-GS rendering speed by reducing redundant sorting operations and preserving rasterization efficiency. GS-TG addresses a critical trade-off issue in 3D-GS rendering: increasing the tile size effectively reduces redundant sorting operations, but it concurrently increases unnecessary rasterization computations. So, during sorting of the proposed approach, GS-TG groups small tiles (for making large tiles) to share sorting operations across tiles within each group, significantly reducing redundant computations. During rasterization, a bitmask assigned to each Gaussian identifies relevant small tiles, to enable efficient sharing of sorting results. Consequently, GS-TG enables sorting to be performed as if a large tile size is used by grouping tiles during the sorting stage, while allowing rasterization to proceed with the original small tiles by using bitmasks in the rasterization stage. GS-TG is a lossless method requiring no retraining or fine-tuning and it can be seamlessly integrated with previous 3D-GS optimization techniques. Experimental results show that GS-TG achieves an average speed-up of 1.54 times over state-of-the-art 3D-GS accelerators.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 15:53:20 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 01:05:48 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Jo",
"Joongho",
""
],
[
"Park",
"Jongsun",
""
]
] |
2509.00930
|
Yanxiao Zhao
|
Yanxiao Zhao, Yaqian Li, Zihao Bo, Rinyoichi Takezoe, Haojia Hui, Mo Guang, Lei Ren, Xiaolin Qin, Kaiwen Long
|
SATQuest: A Verifier for Logical Reasoning Evaluation and Reinforcement Fine-Tuning of LLMs
| null | null | null | null |
cs.AI cs.LG cs.LO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent advances in Large Language Models (LLMs) have demonstrated remarkable general reasoning capabilities. However, systematically evaluating and enhancing these reasoning capabilities is challenging due to the lack of controllable and scalable tools for fine-grained analysis. Existing benchmarks and datasets often lack the necessary variable control for multi-dimensional, systematic analysis and training, or have narrow problem types and formats. To address these limitations, we introduce SATQuest, a systematic verifier designed to evaluate and enhance logical reasoning in LLMs by generating diverse, Satisfiability-based logical reasoning problems directly from Conjunctive Normal Form (CNF) instances. SATQuest structures these problems along three orthogonal dimensions: instance scale, problem type, and question format, employing randomized, SAT-based problem generation and objective answer verification via PySAT. This design mitigates memorization issues, allows for nuanced insights into reasoning performance, and enables effective reinforcement fine-tuning. Our extensive evaluation of various LLMs using SATQuest identified significant limitations in their logical reasoning, particularly in generalizing beyond familiar mathematical formats. Furthermore, we show that reinforcement fine-tuning with SATQuest rewards substantially improves targeted task performance and generalizes to more complex instances, while highlighting remaining challenges in cross-format adaptation. Through these demonstrations, we showcase SATQuest's potential as a foundational tool and a valuable starting point for advancing LLM logical reasoning.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 16:56:06 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Zhao",
"Yanxiao",
""
],
[
"Li",
"Yaqian",
""
],
[
"Bo",
"Zihao",
""
],
[
"Takezoe",
"Rinyoichi",
""
],
[
"Hui",
"Haojia",
""
],
[
"Guang",
"Mo",
""
],
[
"Ren",
"Lei",
""
],
[
"Qin",
"Xiaolin",
""
],
[
"Long",
"Kaiwen",
""
]
] |
2509.00971
|
Vishvesh Bhat
|
Jay Vaghasiya, Omkar Ghugarkar, Vishvesh Bhat, Vipul Dholaria, Julian McAuley
|
CoreThink: A Symbolic Reasoning Layer to reason over Long Horizon Tasks with LLMs
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We introduce CoreThink, a state-of-the-art Reasoning Layer built upon a novel reasoning method called General Symbolics. This approach diverges from reasoning paradigms such as test-time scaling, Supervised Fine-Tuning (SFT), and Reinforcement Learning with Verifiable Rewards (RLVR). CoreThink General Symbolic Reasoner (GSR) is specifically structured around three key use cases: tool-calling, code generation, and planning, demonstrating exemplary performance across a total of seven benchmarks in their respective areas. Notably, we are achieving SOTA scores of 66.66% on Livecodebench v6, 89% on Instruction-Following Evals, and 24.4% on ARC-AGI-2. We also present an agentic coding IDE, developed using the principles of General Symbolics, which achieves a state-of-the-art accuracy of 62.3% on SWE-Bench Lite. We are able to achieve these improvements without any fine-tuning or training costs. Our Reasoning Layer is designed to provide a pure performance uplift, ensuring that a model's accuracy on reasoning tasks is never negatively impacted. We argue that incumbent methods will eventually lead to diminishing returns in LLM performance, necessitating the development of new reasoning techniques. This technical report details our approach at a high level and the availability of the CoreThink models for reasoning-intensive use cases.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 19:31:53 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 05:36:42 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Vaghasiya",
"Jay",
""
],
[
"Ghugarkar",
"Omkar",
""
],
[
"Bhat",
"Vishvesh",
""
],
[
"Dholaria",
"Vipul",
""
],
[
"McAuley",
"Julian",
""
]
] |
2509.01005
|
Jes\'us Oliva-Maza
|
J. Oliva-Maza, Y. Tomilov
|
On similarity to contraction semigroups and tensor products, I
|
36 pages
| null | null | null |
math.FA
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
In the context of finite tensor products of Hilbert spaces, we prove that similarity of a tensor product of operator semigroups to a contraction semigroup is equivalent to the corresponding similarity for each factor, after an appropriate rescaling. A similar result holds with contractivity replaced by quasi-contractivity. This splitting phenomenon allows us to construct new and, in a sense, the strongest possible examples of $C_0$-semigroups not similar to contractions, thus completing an important chapter of the theory. We also address the discrete setting and relate it to our results.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 21:55:36 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 10:41:47 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Oliva-Maza",
"J.",
""
],
[
"Tomilov",
"Y.",
""
]
] |
2509.01007
|
Jes\'us Oliva-Maza
|
J. Oliva-Maza and Y. Tomilov
|
Similarity to contraction semigroups: structural properties, criteria, and applications to control theory
|
73 pages
| null | null | null |
math.FA
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
We reveal new aspects of the structure of Hilbert space $C_0$-semigroups $\mathcal T = (T(t))_{t\ge 0}$ similar to semigroups of contractions. In particular, we prove that $\mathcal T$ is similar to a semigroup of contractions if and only if $\mathcal T$ is similar to a quasi-contraction $C_0$-semigroup and $T(t)$ is similar to a contraction for a single $t>0.$ Moreover, our methods allow us to estimate the corresponding similarity constants and clarify their role in the study of similarity to contractions. Along the way, we obtain similarity conditions involving unbounded operators and imposing minimal assumptions on regularity of $\mathcal T$. Such a general setting allows us to find significant applications to control theory, including characterizations of exactly observable and exactly controllable systems. Finally we establish several criteria of the same flavor for similarity to isometric semigroups, and illustrate the developed theory by a number of pertinent examples.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 21:57:39 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 10:39:44 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Oliva-Maza",
"J.",
""
],
[
"Tomilov",
"Y.",
""
]
] |
2509.01028
|
Zixin Zhu
|
Zixin Zhu, Kevin Duarte, Mamshad Nayeem Rizve, Chengyuan Xu, Ratheesh Kalarot, Junsong Yuan
|
CompSlider: Compositional Slider for Disentangled Multiple-Attribute Image Generation
|
Accepted by ICCV 2025
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
In text-to-image (T2I) generation, achieving fine-grained control over attributes - such as age or smile - remains challenging, even with detailed text prompts. Slider-based methods offer a solution for precise control of image attributes. Existing approaches typically train individual adapter for each attribute separately, overlooking the entanglement among multiple attributes. As a result, interference occurs among different attributes, preventing precise control of multiple attributes together. To address this challenge, we aim to disentangle multiple attributes in slider-based generation to enbale more reliable and independent attribute manipulation. Our approach, CompSlider, can generate a conditional prior for the T2I foundation model to control multiple attributes simultaneously. Furthermore, we introduce novel disentanglement and structure losses to compose multiple attribute changes while maintaining structural consistency within the image. Since CompSlider operates in the latent space of the conditional prior and does not require retraining the foundation model, it reduces the computational burden for both training and inference. We evaluate our approach on a variety of image attributes and highlight its generality by extending to video generation.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 23:36:44 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Sep 2025 15:01:47 GMT"
}
] | 2025-09-04T00:00:00 |
[
[
"Zhu",
"Zixin",
""
],
[
"Duarte",
"Kevin",
""
],
[
"Rizve",
"Mamshad Nayeem",
""
],
[
"Xu",
"Chengyuan",
""
],
[
"Kalarot",
"Ratheesh",
""
],
[
"Yuan",
"Junsong",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.