id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2509.03629
|
Donglin Zhu
|
Donglin Zhu, Peiyao Li, and Ge Jin
|
Noise is All You Need: rethinking the value of noise on seismic denoising via diffusion models
| null | null | null | null |
physics.geo-ph
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
We introduce SeisDiff-denoNIA, a novel diffusion-based seismic denoising framework that trains directly on field noise, eliminating the reliance on synthetic datasets. Unlike conventional denoising methods that require clean signal labels, our approach leverages field noise extracted prior to first arrivals as training targets, allowing the diffusion model to explicitly learn the true noise distribution. The model demonstrates robust performance on field DAS-VSP data contaminated by different noise types, and significantly outperforms traditional signal-based diffusion models under low SNR conditions in synthetic tests. The results suggest that explicitly modeling noise is not only viable but advantageous for seismic denoising tasks.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 18:30:48 GMT"
},
{
"version": "v2",
"created": "Fri, 05 Sep 2025 21:19:02 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhu",
"Donglin",
""
],
[
"Li",
"Peiyao",
""
],
[
"Jin",
"Ge",
""
]
] |
2509.03646
|
Haozhe Wang
|
Haozhe Wang, Qixin Xu, Che Liu, Junhong Wu, Fangzhen Lin, Wenhu Chen
|
Emergent Hierarchical Reasoning in LLMs through Reinforcement Learning
|
Preprint
| null | null | null |
cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Reinforcement Learning (RL) has proven highly effective at enhancing the complex reasoning abilities of Large Language Models (LLMs), yet underlying mechanisms driving this success remain largely opaque. Our analysis reveals that puzzling phenomena like ``aha moments", ``length-scaling'' and entropy dynamics are not disparate occurrences but hallmarks of an emergent reasoning hierarchy, akin to the separation of high-level strategic planning from low-level procedural execution in human cognition. We uncover a compelling two-phase dynamic: initially, a model is constrained by procedural correctness and must improve its low-level skills. The learning bottleneck then decisively shifts, with performance gains being driven by the exploration and mastery of high-level strategic planning. This insight exposes a core inefficiency in prevailing RL algorithms like GRPO, which apply optimization pressure agnostically and dilute the learning signal across all tokens. To address this, we propose HIerarchy-Aware Credit Assignment (HICRA), an algorithm that concentrates optimization efforts on high-impact planning tokens. HICRA significantly outperforms strong baselines, demonstrating that focusing on this strategic bottleneck is key to unlocking advanced reasoning. Furthermore, we validate semantic entropy as a superior compass for measuring strategic exploration over misleading metrics such as token-level entropy.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 18:52:49 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 11:03:11 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wang",
"Haozhe",
""
],
[
"Xu",
"Qixin",
""
],
[
"Liu",
"Che",
""
],
[
"Wu",
"Junhong",
""
],
[
"Lin",
"Fangzhen",
""
],
[
"Chen",
"Wenhu",
""
]
] |
2509.03652
|
Arshak Hovhannisyan
|
E. Khalafyan, A. E. Allahverdyan and A. Hovhannisyan
|
Nonnegative matrix factorization and the principle of the common cause
| null | null | null | null |
cs.LG cs.AI physics.data-an stat.ML
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Nonnegative matrix factorization (NMF) is a known unsupervised data-reduction method. The principle of the common cause (PCC) is a basic methodological approach in probabilistic causality, which seeks an independent mixture model for the joint probability of two dependent random variables. It turns out that these two concepts are closely related. This relationship is explored reciprocally for several datasets of gray-scale images, which are conveniently mapped into probability models. On one hand, PCC provides a predictability tool that leads to a robust estimation of the effective rank of NMF. Unlike other estimates (e.g., those based on the Bayesian Information Criteria), our estimate of the rank is stable against weak noise. We show that NMF implemented around this rank produces features (basis images) that are also stable against noise and against seeds of local optimization, thereby effectively resolving the NMF nonidentifiability problem. On the other hand, NMF provides an interesting possibility of implementing PCC in an approximate way, where larger and positively correlated joint probabilities tend to be explained better via the independent mixture model. We work out a clustering method, where data points with the same common cause are grouped into the same cluster. We also show how NMF can be employed for data denoising.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 19:02:39 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Khalafyan",
"E.",
""
],
[
"Allahverdyan",
"A. E.",
""
],
[
"Hovhannisyan",
"A.",
""
]
] |
2509.03654
|
Edgardo Ugalde
|
Andrea Espa\~na, William Funez and Edgardo Ugalde
|
Dominant vertices and attractors' landscape for Boolean networks
| null | null | null | null |
math.DS nlin.CG q-bio.MN
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In previous works, we introduced the notion of dominant vertices. This is a set of nodes in the underlying network whose evolution determines the whole network's dynamics after a transient time. In this paper, we focus on the case of Boolean networks. We define a reduced graph on the dominant vertices and an induced automata network on this graph, which we prove is asymptotically equivalent to the original Boolean dynamics. Asymptotic conjugacy ensures that the systems, restricted to their respective attractors, are dynamically equivalent. For a significant class of networks, the induced automata network is indeed a reduction of the original system. In these cases, the reduction, which is obtained from the structure of dominant vertices, supplies a more tractable system with the same structure of attractors as the original one. Furthermore, the structure of the induced system allows us to establish bounds on the number and period of the attractors, as well as on the reduction of the basin's sizes and transient lengths. We illustrate this reduction by considering a class of networks, which we call clover networks, whose dominant set is a singleton. To get insight into the structure of the basins of attraction of Boolean networks with a single dominant vertex, we complement this work with a numerical exploration of the behavior of a parametrized ensemble of systems of this kind.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 19:03:26 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 02:22:48 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"España",
"Andrea",
""
],
[
"Funez",
"William",
""
],
[
"Ugalde",
"Edgardo",
""
]
] |
2509.03658
|
Antonio Guillen-Perez
|
Antonio Guillen-Perez
|
Efficient Virtuoso: A Latent Diffusion Transformer Model for Goal-Conditioned Trajectory Planning
| null | null | null | null |
cs.RO cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The ability to generate a diverse and plausible distribution of future trajectories is a critical capability for autonomous vehicle planning systems. While recent generative models have shown promise, achieving high fidelity, computational efficiency, and precise control remains a significant challenge. In this paper, we present the Efficient Virtuoso, a conditional latent diffusion model for goal-conditioned trajectory planning. Our approach introduces a novel two-stage normalization pipeline that first scales trajectories to preserve their geometric aspect ratio and then normalizes the resulting PCA latent space to ensure a stable training target. The denoising process is performed efficiently in this low-dimensional latent space by a simple MLP denoiser, which is conditioned on a rich scene context fused by a powerful Transformer-based StateEncoder. We demonstrate that our method achieves state-of-the-art performance on the Waymo Open Motion Dataset, achieving a minimum Average Displacement Error (minADE) of 0.25. Furthermore, through a rigorous ablation study on goal representation, we provide a key insight: while a single endpoint goal can resolve strategic ambiguity, a richer, multi-step sparse route is essential for enabling the precise, high-fidelity tactical execution that mirrors nuanced human driving behavior.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 19:18:02 GMT"
},
{
"version": "v2",
"created": "Sat, 06 Sep 2025 15:10:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Guillen-Perez",
"Antonio",
""
]
] |
2509.03663
|
Rudi Brits
|
Rudi Brits, Muhammad Hassen and Cheick Toure
|
The Multiplicative Kowalski-Slodkowski Theorem for Hermitian Algebras
|
7 pages
| null | null | null |
math.FA
|
http://creativecommons.org/licenses/by/4.0/
|
We prove, for Hermitian algebras, the multiplicative version of the Kowalski-S\l{}odkowski Theorem which identifies the characters among the collection of all complex valued functions on a Banach algebra $A$ in terms of a spectral condition. Specifically, we show that, if $A$ is a Hermitian algebra, and if $\phi:A\mapsto\mathbb C$ is a continuous function satisfying $\phi(x)\phi(y) \in \sigma(xy)$ for all $x,y\in A$ (where $\sigma$ denotes the spectrum), then either $\phi$ or $-\phi$ is a character of $A$; of course the converse holds as well. Our proof depends fundamentally on the existence of positive elements and square roots in these algebras.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 19:26:59 GMT"
},
{
"version": "v2",
"created": "Sat, 06 Sep 2025 11:00:49 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Brits",
"Rudi",
""
],
[
"Hassen",
"Muhammad",
""
],
[
"Toure",
"Cheick",
""
]
] |
2509.03677
|
Daniel Yun
|
Vincent-Daniel Yun
|
Insights from Gradient Dynamics: Gradient Autoscaled Normalization
| null | null | null | null |
cs.LG cs.AI cs.CV cs.IT math.IT
|
http://creativecommons.org/publicdomain/zero/1.0/
|
Gradient dynamics play a central role in determining the stability and generalization of deep neural networks. In this work, we provide an empirical analysis of how variance and standard deviation of gradients evolve during training, showing consistent changes across layers and at the global scale in convolutional networks. Motivated by these observations, we propose a hyperparameter-free gradient normalization method that aligns gradient scaling with their natural evolution. This approach prevents unintended amplification, stabilizes optimization, and preserves convergence guarantees. Experiments on the challenging CIFAR-100 benchmark with ResNet-20, ResNet-56, and VGG-16-BN demonstrate that our method maintains or improves test accuracy even under strong generalization. Beyond practical performance, our study highlights the importance of directly tracking gradient dynamics, aiming to bridge the gap between theoretical expectations and empirical behaviors, and to provide insights for future optimization research.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 19:54:23 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 06:17:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Yun",
"Vincent-Daniel",
""
]
] |
2509.03724
|
Rob Spaargaren
|
Rob J. Spaargaren, Oliver Herbort, Haiyang S. Wang, Stephen J. Mojzsis, Paolo Sossi
|
Proto-planetary disk composition-dependent element volatility in the context of rocky planet formation
|
19 pages, 16 figures; Accepted by A&A on 05/09/2025
| null | null | null |
astro-ph.EP
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
The compositions of the Solar System terrestrial bodies are fractionated from that of the Sun, where elemental depletions in the bulk rocky bodies correlate with element volatility, expressed in its 50% condensation temperature. However, because element volatility depends on disk gas composition, it is not mandated that elemental fractionation trends derived from the solar-terrestrial scenario apply to other planetary systems. Here, we expand upon previous efforts to quantify element volatility during disk condensation, and how this affects rocky planet compositional diversity.
We simulate condensation sequences for a sample of 1000 initial disk compositions based on observed stellar abundances. We present parametrisations of how element 50% condensation temperatures depend on disk composition, and apply element fractionation trends with appropriate element volatilty to stellar abundances to simulate compositions of rocky exoplanets with the same volatile depletion pattern as the Earth, providing a robust and conservative lower limit to the compositional diversity of rocky exoplanets.
Here we show that Earth-like planets emerge from low-C-to-O disks and graphite-bearing planets from medium-to-high-C-to-O disks. Furthermore, we identify an intermediate-C-to-O class of planets characterized by Mg and Si depletion, leading to relatively high abundances of Fe, Ca, and Al. We show that devolatilisation patterns could be adapted potentially with disk composition-dependent condensation temperatures to make predictions of rocky planet bulk compositions within individual systems. The outcomes of our analysis suggest that accounting for disk composition-dependent condensation temperatures means that we can expect an even broader range of possible rocky planet compositions than has previously been considered.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 21:11:47 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 13:57:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Spaargaren",
"Rob J.",
""
],
[
"Herbort",
"Oliver",
""
],
[
"Wang",
"Haiyang S.",
""
],
[
"Mojzsis",
"Stephen J.",
""
],
[
"Sossi",
"Paolo",
""
]
] |
2509.03736
|
James Mooney
|
James Mooney, Josef Woldense, Zheng Robert Jia, Shirley Anugrah Hayati, My Ha Nguyen, Vipul Raheja, Dongyeop Kang
|
Are LLM Agents Behaviorally Coherent? Latent Profiles for Social Simulation
|
25 pages, 9 figures, 7 tables
| null | null | null |
cs.AI cs.CL cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
The impressive capabilities of Large Language Models (LLMs) have fueled the notion that synthetic agents can serve as substitutes for real participants in human-subject research. In an effort to evaluate the merits of this claim, social science researchers have largely focused on whether LLM-generated survey data corresponds to that of a human counterpart whom the LLM is prompted to represent. In contrast, we address a more fundamental question: Do agents maintain internal consistency, retaining similar behaviors when examined under different experimental settings? To this end, we develop a study designed to (a) reveal the agent's internal state and (b) examine agent behavior in a basic dialogue setting. This design enables us to explore a set of behavioral hypotheses to assess whether an agent's conversation behavior is consistent with what we would expect from their revealed internal state. Our findings on these hypotheses show significant internal inconsistencies in LLMs across model families and at differing model sizes. Most importantly, we find that, although agents may generate responses matching those of their human counterparts, they fail to be internally consistent, representing a critical gap in their capabilities to accurately substitute for real participants in human-subject research. Our simulation code and data are publicly accessible.
|
[
{
"version": "v1",
"created": "Wed, 03 Sep 2025 21:55:29 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Mooney",
"James",
""
],
[
"Woldense",
"Josef",
""
],
[
"Jia",
"Zheng Robert",
""
],
[
"Hayati",
"Shirley Anugrah",
""
],
[
"Nguyen",
"My Ha",
""
],
[
"Raheja",
"Vipul",
""
],
[
"Kang",
"Dongyeop",
""
]
] |
2509.03821
|
Rui Zhao
|
Rui Zhao, Muhammad Shoaib, Viet Tung Hoang, and Wajih Ul Hassan
|
Rethinking Tamper-Evident Logging: A High-Performance, Co-Designed Auditing System
|
This paper has been accepted to the ACM Conference on Computer and Communications Security (CCS) 2025
|
Proceedings of the 2025 ACM SIGSAC Conference on Computer and Communications Security
|
10.1145/3719027.3765024
| null |
cs.CR
|
http://creativecommons.org/licenses/by/4.0/
|
Existing tamper-evident logging systems suffer from high overhead and severe data loss in high-load settings, yet only provide coarse-grained tamper detection. Moreover, installing such systems requires recompiling kernel code. To address these challenges, we present Nitro, a high-performance, tamper-evident audit logging system that supports fine-grained detection of log tampering. Even better, our system avoids kernel recompilation by using the eBPF technology. To formally justify the security of Nitro, we provide a new definitional framework for logging systems, and give a practical cryptographic construction meeting this new goal. Unlike prior work that focus only on the cryptographic processing, we codesign the cryptographic part with the pre- and post-processing of the logs to exploit all system-level optimizations. Our evaluations demonstrate Nitro's superior performance, achieving 10X-25X improvements in high-stress conditions and 2X-10X in real-world scenarios while maintaining near-zero data loss. We also provide an advanced variant, Nitro-R that introduces in-kernel log reduction techniques to reduce runtime overhead even further.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:12:40 GMT"
},
{
"version": "v2",
"created": "Fri, 05 Sep 2025 21:13:13 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhao",
"Rui",
""
],
[
"Shoaib",
"Muhammad",
""
],
[
"Hoang",
"Viet Tung",
""
],
[
"Hassan",
"Wajih Ul",
""
]
] |
2509.03832
|
Georgiy Lapin
|
Joseph Jackson, Georgiy Lapin, Jeremy E. Thompson
|
Gravity Well Echo Chamber Modeling With An LLM-Based Confirmation Bias Model
| null | null | null | null |
cs.SI cs.AI cs.CY
|
http://creativecommons.org/licenses/by/4.0/
|
Social media echo chambers play a central role in the spread of misinformation, yet existing models often overlook the influence of individual confirmation bias. An existing model of echo chambers is the "gravity well" model, which creates an analog between echo chambers and spatial gravity wells. We extend this established model by introducing a dynamic confirmation bias variable that adjusts the strength of pull based on a user's susceptibility to belief-reinforcing content. This variable is calculated for each user through comparisons between their posting history and their responses to posts of a wide range of viewpoints.
Incorporating this factor produces a confirmation-bias-integrated gravity well model that more accurately identifies echo chambers and reveals community-level markers of information health. We validated the approach on nineteen Reddit communities, demonstrating improved detection of echo chambers.
Our contribution is a framework for systematically capturing the role of confirmation bias in online group dynamics, enabling more effective identification of echo chambers. By flagging these high-risk environments, the model supports efforts to curb the spread of misinformation at its most common points of amplification.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 02:41:35 GMT"
},
{
"version": "v2",
"created": "Sun, 07 Sep 2025 00:29:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Jackson",
"Joseph",
""
],
[
"Lapin",
"Georgiy",
""
],
[
"Thompson",
"Jeremy E.",
""
]
] |
2509.03859
|
Haichao Zhang
|
Haichao Zhang, Haonan Yu, Le Zhao, Andrew Choi, Qinxun Bai, Yiqing Yang, Wei Xu
|
Learning Multi-Stage Pick-and-Place with a Legged Mobile Manipulator
|
Accepted to IEEE Robotics and Automation Letters (RA-L). Tech Report: arXiv:2501.09905
| null | null | null |
cs.RO
|
http://creativecommons.org/licenses/by/4.0/
|
Quadruped-based mobile manipulation presents significant challenges in robotics due to the diversity of required skills, the extended task horizon, and partial observability. After presenting a multi-stage pick-and-place task as a succinct yet sufficiently rich setup that captures key desiderata for quadruped-based mobile manipulation, we propose an approach that can train a visuo-motor policy entirely in simulation, and achieve nearly 80\% success in the real world. The policy efficiently performs search, approach, grasp, transport, and drop into actions, with emerged behaviors such as re-grasping and task chaining. We conduct an extensive set of real-world experiments with ablation studies highlighting key techniques for efficient training and effective sim-to-real transfer. Additional experiments demonstrate deployment across a variety of indoor and outdoor environments. Demo videos and additional resources are available on the project page: https://horizonrobotics.github.io/gail/SLIM.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 03:36:07 GMT"
},
{
"version": "v2",
"created": "Fri, 05 Sep 2025 07:29:42 GMT"
},
{
"version": "v3",
"created": "Mon, 08 Sep 2025 04:39:16 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhang",
"Haichao",
""
],
[
"Yu",
"Haonan",
""
],
[
"Zhao",
"Le",
""
],
[
"Choi",
"Andrew",
""
],
[
"Bai",
"Qinxun",
""
],
[
"Yang",
"Yiqing",
""
],
[
"Xu",
"Wei",
""
]
] |
2509.03990
|
Chunlong Wu
|
Chunlong Wu and Ye Luo and Zhibo Qu and Min Wang
|
Meta-Policy Reflexion: Reusable Reflective Memory and Rule Admissibility for Resource-Efficient LLM Agent
| null | null | null | null |
cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large language model (LLM) agents achieve impressive single-task performance but commonly exhibit repeated failures, inefficient exploration, and limited cross-task adaptability. Existing reflective strategies (e.g., Reflexion, ReAct) improve per-episode behavior but typically produce ephemeral, task-specific traces that are not reused across tasks. Reinforcement-learning based alternatives can produce transferable policies but require substantial parameter updates and compute. In this work we introduce Meta-Policy Reflexion (MPR): a hybrid framework that consolidates LLM-generated reflections into a structured, predicate-like Meta-Policy Memory (MPM) and applies that memory at inference time through two complementary mechanisms soft memory-guided decoding and hard rule admissibility checks(HAC). MPR (i) externalizes reusable corrective knowledge without model weight updates, (ii) enforces domain constraints to reduce unsafe or invalid actions, and (iii) retains the adaptability of language-based reflection. We formalize the MPM representation, present algorithms for update and decoding, and validate the approach in a text-based agent environment following the experimental protocol described in the provided implementation (AlfWorld-based). Empirical results reported in the supplied material indicate consistent gains in execution accuracy and robustness when compared to Reflexion baselines; rule admissibility further improves stability. We analyze mechanisms that explain these gains, discuss scalability and failure modes, and outline future directions for multimodal and multi-agent extensions.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 08:18:39 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 07:40:58 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wu",
"Chunlong",
""
],
[
"Luo",
"Ye",
""
],
[
"Qu",
"Zhibo",
""
],
[
"Wang",
"Min",
""
]
] |
2509.04001
|
Xing-Hui Feng
|
Xing-Hui Feng and Guang-Yu Zhang
|
Shadow and Quasi-Normal Modes of Schwarzschild-Hernquist Black Hole
|
18 pages, 3 figures, 2 tables; references updated
| null | null | null |
gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper we study the shadow and quasi-normal modes (QNMs) of a black hole (BH) surrounded by a dark matter halo with Hernquist-type density distribution, which was reported in Ref. \cite{Cardoso:2021wlq}. In astrophysical scenarios, we find that the shadow radius enlarges as the compactness of halo increases. Therefore, we obtain an upper bound for the compactness ${\cal C}\le0.092$ with the Event Horizon Telescope (EHT) observations. We calculate axial gravitational QNMs of the galactic BH up to ${\cal C}\sim{\cal O}(1)$, and fit the redshift relative to Schwarzschild QNMs up to second order in the compactness (for ${\cal C}\le0.3)$. These highly redshifted QNMs, resulting from large compactness, are key to modeling the dark matter halo.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 08:30:33 GMT"
},
{
"version": "v2",
"created": "Sun, 07 Sep 2025 08:58:18 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Feng",
"Xing-Hui",
""
],
[
"Zhang",
"Guang-Yu",
""
]
] |
2509.04015
|
Yf Sun
|
Yu-Fei Sun, Yue Mao, Yu-Chen Zhuang and Qing-Feng Sun
|
Tunneling Magnetoresistance Effect in Altermagnets
|
13 pages, 7 figures
|
Phys. Rev. B 112, 094411 (2025)
|
10.1103/t8b5-l859
| null |
cond-mat.mes-hall
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
As an unconventional magnet, altermagnetism attracts great interest in condensed matter physics and applies a new research platform for the spintronics. Since the tunneling magnetoresistance (TMR) effect is an important research aspect in spintronics, we theoretically propose a universal altermagnetic sandwich device to achieve the TMR effect and investigate its transport properties. Using the nonequilibrium Green's function method and the Landauer-B\"uttiker formula, we obtain the conductance and the TMR ratio. By systematically rotating the orientations of the altermagnet and spin, we investigate how the altermagnetic orientations affect the conductance and the TMR ratio, and comprehensively demonstrate the dependence of the conductance and the TMR ratio on a range of parameters in the system. By tuning the altermagnetism strength and the Fermi energy, as well as rotating the orientations in the altermagnet, the TMR ratio can reach a value of over 1000%. In addition, we analyze the detailed symmetry relations of the conductance and the TMR ratio in our system. Our approach provides a new design concept for the next-generation information technologies based on the altermagnetic platform, paving the way for the development of spintronics applications.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 08:44:28 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Sun",
"Yu-Fei",
""
],
[
"Mao",
"Yue",
""
],
[
"Zhuang",
"Yu-Chen",
""
],
[
"Sun",
"Qing-Feng",
""
]
] |
2509.04039
|
Seong-Hoon Jang Dr.
|
Seong-Hoon Jang and Di Zhang and Hung Ba Tran and Xue Jia and Kiyoe Konno and Ryuhei Sato and Shin-ichi Orimo and Hao Li
|
Physically Interpretable Descriptors Drive the Materials Design of Metal Hydrides for Hydrogen Storage
| null | null | null | null |
cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Designing metal hydrides for hydrogen storage remains a longstanding challenge due to the vast compositional space and complex structure-property relationships. Herein, for the first time, we present physically interpretable models for predicting two key performance metrics, gravimetric hydrogen density $w$ and equilibrium pressure $P_{\rm eq,RT}$ at room temperature, based on a minimal set of chemically meaningful descriptors. Using a rigorously curated dataset of $5,089$ metal hydride compositions from our recently developed Digital Hydrogen Platform (\it{DigHyd}) based on large-scale data mining from available experimental literature of solid-state hydrogen storage materials, we systematically constructed over $1.6$ million candidate models using combinations of scalar transformations and nonlinear link functions. The final closed-form models, derived from $2$-$3$ descriptors each, achieve predictive accuracies on par with state-of-the-art machine learning methods, while maintaining full physical transparency. Strikingly, descriptor-based design maps generated from these models reveal a fundamental trade-off between $w$ and $P_{\rm eq,RT}$: saline-type hydrides, composed of light electropositive elements, offer high $w$ but low $P_{\rm eq,RT}$, whereas interstitial-type hydrides based on heavier electronegative transition metals show the opposite trend. Notably, Be-based systems, such as Be-Na alloys, emerge as rare candidates that simultaneously satisfy both performance metrics, attributed to the unique combination of light mass and high molar density for Be. Our models indicate that Be-based systems may offer renewed prospects for approaching these benchmarks. These results provide chemically intuitive guidelines for materials design and establish a scalable framework for the rational discovery of materials in complex chemical spaces.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 09:18:04 GMT"
},
{
"version": "v2",
"created": "Sat, 06 Sep 2025 19:46:37 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Jang",
"Seong-Hoon",
""
],
[
"Zhang",
"Di",
""
],
[
"Tran",
"Hung Ba",
""
],
[
"Jia",
"Xue",
""
],
[
"Konno",
"Kiyoe",
""
],
[
"Sato",
"Ryuhei",
""
],
[
"Orimo",
"Shin-ichi",
""
],
[
"Li",
"Hao",
""
]
] |
2509.04078
|
Liu Jingjing
|
Jingjing Liu, Zeming Liu, Zihao Cheng, Mengliang He, Xiaoming Shi, Yuhang Guo, Xiangrong Zhu, Yuanfang Guo, Yunhong Wang, Haifeng Wang
|
RepoDebug: Repository-Level Multi-Task and Multi-Language Debugging Evaluation of Large Language Models
|
30 pages, 12 figures, EMNLP 2025 Findings
| null | null | null |
cs.SE cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Large Language Models (LLMs) have exhibited significant proficiency in code debugging, especially in automatic program repair, which may substantially reduce the time consumption of developers and enhance their efficiency. Significant advancements in debugging datasets have been made to promote the development of code debugging. However, these datasets primarily focus on assessing the LLM's function-level code repair capabilities, neglecting the more complex and realistic repository-level scenarios, which leads to an incomplete understanding of the LLM's challenges in repository-level debugging. While several repository-level datasets have been proposed, they often suffer from limitations such as limited diversity of tasks, languages, and error types. To mitigate this challenge, this paper introduces RepoDebug, a multi-task and multi-language repository-level code debugging dataset with 22 subtypes of errors that supports 8 commonly used programming languages and 3 debugging tasks. Furthermore, we conduct evaluation experiments on 10 LLMs, where Claude 3.5 Sonnect, the best-performing model, still cannot perform well in repository-level debugging.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 10:13:21 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 08:22:38 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Liu",
"Jingjing",
""
],
[
"Liu",
"Zeming",
""
],
[
"Cheng",
"Zihao",
""
],
[
"He",
"Mengliang",
""
],
[
"Shi",
"Xiaoming",
""
],
[
"Guo",
"Yuhang",
""
],
[
"Zhu",
"Xiangrong",
""
],
[
"Guo",
"Yuanfang",
""
],
[
"Wang",
"Yunhong",
""
],
[
"Wang",
"Haifeng",
""
]
] |
2509.04144
|
Malte Londschien
|
Malte Londschien
|
The exact distribution of the conditional likelihood-ratio test in instrumental variables regression
| null | null | null | null |
econ.EM math.ST stat.TH
|
http://creativecommons.org/licenses/by/4.0/
|
We derive the exact asymptotic distribution of the conditional likelihood-ratio test in instrumental variables regression under weak instrument asymptotics and for multiple endogenous variables. The distribution is conditional on all eigenvalues of the concentration matrix, rather than only the smallest eigenvalue as in an existing asymptotic upper bound. This exact characterization leads to a substantially more powerful test if there are differently identified endogenous variables. We provide computational methods implementing the test and demonstrate the power gains through numerical analysis.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 12:15:25 GMT"
},
{
"version": "v2",
"created": "Sun, 07 Sep 2025 12:16:33 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Londschien",
"Malte",
""
]
] |
2509.04145
|
Dongliang Cao
|
Dongliang Cao, Guoxing Sun, Marc Habermann, Florian Bernard
|
Hyper Diffusion Avatars: Dynamic Human Avatar Generation using Network Weight Space Diffusion
|
Project webpage: https://vcai.mpi-inf.mpg.de/projects/HDA/
| null | null | null |
cs.GR cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Creating human avatars is a highly desirable yet challenging task. Recent advancements in radiance field rendering have achieved unprecedented photorealism and real-time performance for personalized dynamic human avatars. However, these approaches are typically limited to person-specific rendering models trained on multi-view video data for a single individual, limiting their ability to generalize across different identities. On the other hand, generative approaches leveraging prior knowledge from pre-trained 2D diffusion models can produce cartoonish, static human avatars, which are animated through simple skeleton-based articulation. Therefore, the avatars generated by these methods suffer from lower rendering quality compared to person-specific rendering methods and fail to capture pose-dependent deformations such as cloth wrinkles. In this paper, we propose a novel approach that unites the strengths of person-specific rendering and diffusion-based generative modeling to enable dynamic human avatar generation with both high photorealism and realistic pose-dependent deformations. Our method follows a two-stage pipeline: first, we optimize a set of person-specific UNets, with each network representing a dynamic human avatar that captures intricate pose-dependent deformations. In the second stage, we train a hyper diffusion model over the optimized network weights. During inference, our method generates network weights for real-time, controllable rendering of dynamic human avatars. Using a large-scale, cross-identity, multi-view video dataset, we demonstrate that our approach outperforms state-of-the-art human avatar generation methods.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 12:15:55 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 08:40:58 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Cao",
"Dongliang",
""
],
[
"Sun",
"Guoxing",
""
],
[
"Habermann",
"Marc",
""
],
[
"Bernard",
"Florian",
""
]
] |
2509.04199
|
Dieter Schwarzmann
|
Dieter Schwarzmann and Simon K\"aser
|
On the Effect of Sampling-Time Jitter
|
Updated Version of the one submitted. Submitted for review as letter in IEEE Journal for Transactions on Control Systems Technology
| null | null | null |
eess.SY cs.SY eess.SP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This brief, aimed at practitioners, offers an analysis of the effect of sampling-time jitter, i. e., the error produced by execution-time inaccuracies. We propose reinterpreting jitter-afflicted linear time-invariant systems through equivalent jitter-free analogs. By constructing a perceived system that absorbs the effects of timing perturbations into its dynamics, we find an affine scaling of jitter. We examine both measurement and implementation scenarios, demonstrating that the presence of jitter effectively scales the system matrices. Moreover, we observe that, in the Laplace domain, jitter can be interpreted as a frequency scaling.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 13:23:00 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 16:46:16 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Schwarzmann",
"Dieter",
""
],
[
"Käser",
"Simon",
""
]
] |
2509.04200
|
Martin Hampenberg
|
Martin Hampenberg
|
Submonoids of Infinite Symmetric Inverse Monoids
|
This is a PhD thesis submitted January 2025, 125 pages, 17 figures
| null | null | null |
math.RA math.CO math.GN math.GR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this thesis we study the subsemigroup structure of the symmetric inverse monoid $I_X$, the inverse semigroup of bijections between subsets of the set $X$, when $X$ is an infinite set. We explore three different approaches to this task. First, we classify the maximal subsemigroups of $I_X$ containing certain subgroups of the symmetric group on $X$. The subgroups in question are the symmetric group itself, the pointwise stabiliser of a finite non-empty subset of $X$, the stabiliser of an ultrafilter on $X$, and the stabiliser of a finite partition of $X$. Next, we study subsemigroups of $I_X$ which are closed in semigroup topologies on $I_X$ introduced by Elliot et al. in 2023. We discover that the closed subsemigroups in these topologies that contain all the idempotents of $I_X$ coincide exactly with semigroups of partial endomorphisms and partial automorphisms of relational structures defined on $X$. Furthermore, we show that if a relational structure $R$ on a countable set $X$ only contains a finite number of relations, then there exists a finite subset $U$ of $I_X$ such that the union of the partial automorphisms of $R$ together with $U$ generates all of $I_X$. Finally, we study the subsemigroup structure of $I_X$ under a preorder introduced by George Bergman and Saharon Shelah in 2006 for the symmetric group. Extending the preorder to $I_X$, if $S_1$ and $S_2$ are subsemigroups of $I_X$, we say that $S_1 \preccurlyeq S_2$ if there exists a finite subset $U$ of $I_X$ such that $S_1$ is contained in the semigroup generated by the union of $S_2$ and $U$. We classify certain types of subsemigroups of $I_X$ according the Bergman-Shelah preorder, and we formulate a conjecture analogous to the main result by Bergman and Shelah.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 13:24:51 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 12:17:11 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hampenberg",
"Martin",
""
]
] |
2509.04242
|
Rapha\"el Maire
|
Rapha\"el Maire and Ludivine Chaix
|
Hyperuniformity and conservation laws in non-equilibrium systems
|
26 pages, 6 figures, updated non-linearity section
| null | null | null |
cond-mat.stat-mech cond-mat.soft
|
http://creativecommons.org/licenses/by/4.0/
|
We demonstrate that hyperuniformity, the suppression of density fluctuations at large length scales, emerges generically from the interplay between conservation laws and non-equilibrium driving. The underlying mechanism for this emergence is analogous to self-organized criticality. Based on this understanding, we introduce four non-equilibrium models that consistently demonstrate hyperuniformity. Furthermore, we show that systems with an arbitrary number of conserved mass multipole moments exhibit an arbitrary strong tunable hyperuniform scaling, with the structure factor following $S(k) \sim k^m$, where $m$ is set by the number of conserved multipoles. Finally, we find that hyperuniformity arising from a combination of conserved noise and partially conserved average motion is not robust against non-linear perturbations. These results highlight the central role of conservation laws in stabilizing hyperuniformity and reveal a unifying mechanism for its emergence in non-equilibrium systems.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 14:16:44 GMT"
},
{
"version": "v2",
"created": "Sun, 07 Sep 2025 10:57:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Maire",
"Raphaël",
""
],
[
"Chaix",
"Ludivine",
""
]
] |
2509.04279
|
Priyansh Lunia
|
P. Lunia, A.O. Nelson, C. Paz-Soldan
|
Energy Confinement Time Scaling Law Derived from Paz-Soldan NF 2024
|
Updated language and clarity
| null | null | null |
physics.plasm-ph
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Results from C. Paz-Soldan et al 2024 Nucl. Fusion 64 094002 demonstrate encouraging energy confinement properties for the negative triangularity scenario, similar to or exceeding the scaling of the IPB98(y,2) law. This work describes the procedure with which a new scaling law was regressed specifically from the data from the aforementioned article. Given the relatively small size of the single-machine dataset, kernel density estimation was employed to minimize sampling bias and bootstrapping was used to give a realistic estimate of the large uncertainties from the regression. The resulting power law shows a robustly stronger dependence on plasma current and more severe power degradation as compared to the H-mode scaling law.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 14:55:46 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 14:12:09 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lunia",
"P.",
""
],
[
"Nelson",
"A. O.",
""
],
[
"Paz-Soldan",
"C.",
""
]
] |
2509.04288
|
Rudi Coppola
|
Rudi Coppola, Hovsep Touloujian, Pierfrancesco Ombrini, Manuel Mazo Jr
|
Reinforcement Learning for Robust Ageing-Aware Control of Li-ion Battery Systems with Data-Driven Formal Verification
| null | null | null | null |
eess.SY cs.AI cs.SY
|
http://creativecommons.org/licenses/by/4.0/
|
Rechargeable lithium-ion (Li-ion) batteries are a ubiquitous element of modern technology. In the last decades, the production and design of such batteries and their adjacent embedded charging and safety protocols, denoted by Battery Management Systems (BMS), has taken central stage. A fundamental challenge to be addressed is the trade-off between the speed of charging and the ageing behavior, resulting in the loss of capacity in the battery cell. We rely on a high-fidelity physics-based battery model and propose an approach to data-driven charging and safety protocol design. Following a Counterexample-Guided Inductive Synthesis scheme, we combine Reinforcement Learning (RL) with recent developments in data-driven formal methods to obtain a hybrid control strategy: RL is used to synthesise the individual controllers, and a data-driven abstraction guides their partitioning into a switched structure, depending on the initial output measurements of the battery. The resulting discrete selection among RL-based controllers, coupled with the continuous battery dynamics, realises a hybrid system. When a design meets the desired criteria, the abstraction provides probabilistic guarantees on the closed-loop performance of the cell.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 15:01:03 GMT"
},
{
"version": "v2",
"created": "Sat, 06 Sep 2025 20:36:57 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Coppola",
"Rudi",
""
],
[
"Touloujian",
"Hovsep",
""
],
[
"Ombrini",
"Pierfrancesco",
""
],
[
"Mazo",
"Manuel",
"Jr"
]
] |
2509.04384
|
Danny Marfatia
|
Jo\~ao Gon\c{c}alves, Danny Marfatia, Ant\'onio P. Morais
|
Prominence: A discriminator of gravitational wave signals
|
20 pages, 12 figures, 1 table
| null | null | null |
hep-ph astro-ph.CO gr-qc hep-ex
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The concept of prominence is familiar to signal engineers, topographers and mountaineers. We introduce Prominence $\mathcal P$ as a discriminator of gravitational wave (GW) signals. We treat black hole and neutron star binaries as astrophysical background sources, and show how $\mathcal P$ can be used to distinguish between GW spectra produced by first-order phase transitions, domain walls and cosmic strings, and combinations thereof. Prominence can also be used to discriminate between these and off-piste sources of GWs. The uncertainty in the measured energy density in GWs at Pulsar Timing Arrays needs to be smaller than $\sim 4\%$ for $\mathcal{P}$ to achieve discrimination at 3$\sigma$. LISA and ET data are expected to have sufficiently small uncertainties that Prominence can play a central role in their analysis.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 16:47:32 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Gonçalves",
"João",
""
],
[
"Marfatia",
"Danny",
""
],
[
"Morais",
"António P.",
""
]
] |
2509.04389
|
Alec Riso
|
Alec L. Riso, Karthik Thyagarajan, Connor Whiting, Katherine Jimenez
|
Constructing a Photonic Implementation of Quantum Key Distribution
| null | null | null | null |
quant-ph physics.ed-ph physics.optics
|
http://creativecommons.org/licenses/by/4.0/
|
Quantum Key Distribution (QKD) stands as a revolutionary approach to secure communication, using the principles of quantum mechanics to establish unbreakable channels. Unlike traditional cryptography, which relies on the computational difficulty of mathematical problems, QKD utilizes the inherent properties of quantum states to achieve information-theoretic security. This means that the security of the key exchange is guaranteed by the laws of physics, making it theoretically unbreakable even by an adversary with unlimited computational power. Currently, one of the most viable ways to implement QKD for communication is via photonics, namely, using phase-preserving long-distance optical fibers. The objective of this project is to implement photonic QKD in a laboratory setting. This will help demonstrate the protocol's robustness and provide a feasible implementation for educational demonstrations.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 16:59:17 GMT"
},
{
"version": "v2",
"created": "Sun, 07 Sep 2025 17:05:51 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Riso",
"Alec L.",
""
],
[
"Thyagarajan",
"Karthik",
""
],
[
"Whiting",
"Connor",
""
],
[
"Jimenez",
"Katherine",
""
]
] |
2509.04439
|
Matthew Ho
|
Matthew Ho, Chen Si, Zhaoxiang Feng, Fangxu Yu, Yichi Yang, Zhijian Liu, Zhiting Hu, Lianhui Qin
|
ArcMemo: Abstract Reasoning Composition with Lifelong LLM Memory
| null | null | null | null |
cs.AI cs.CL cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
While inference-time scaling enables LLMs to carry out increasingly long and capable reasoning traces, the patterns and insights uncovered during these traces are immediately discarded once the context window is reset for a new query. External memory is a natural way to persist these discoveries, and recent work has shown clear benefits for reasoning-intensive tasks. We see an opportunity to make such memories more broadly reusable and scalable by moving beyond instance-based memory entries (e.g. exact query/response pairs, or summaries tightly coupled with the original problem context) toward concept-level memory: reusable, modular abstractions distilled from solution traces and stored in natural language. For future queries, relevant concepts are selectively retrieved and integrated into the prompt, enabling test-time continual learning without weight updates. Our design introduces new strategies for abstracting takeaways from rollouts and retrieving entries for new queries, promoting reuse and allowing memory to expand with additional experiences. We evaluate on ARC-AGI, a benchmark that stresses compositional generalization and abstract reasoning, making it a natural fit for concept memory. Our method yields a 7.5% relative gain over a strong no-memory baseline with performance continuing to scale with inference compute. We find abstract concepts to be the most consistent memory design, outscoring the baseline at all tested inference compute scales. Moreover, dynamically updating memory during test-time outperforms fixed settings, supporting the hypothesis that accumulating and abstracting patterns enables further solutions in a form of self-improvement. Code is available at https://github.com/matt-seb-ho/arc_memo.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 17:54:19 GMT"
},
{
"version": "v2",
"created": "Fri, 05 Sep 2025 18:25:53 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ho",
"Matthew",
""
],
[
"Si",
"Chen",
""
],
[
"Feng",
"Zhaoxiang",
""
],
[
"Yu",
"Fangxu",
""
],
[
"Yang",
"Yichi",
""
],
[
"Liu",
"Zhijian",
""
],
[
"Hu",
"Zhiting",
""
],
[
"Qin",
"Lianhui",
""
]
] |
2509.04441
|
Hao-Shu Fang
|
Hao-Shu Fang, Branden Romero, Yichen Xie, Arthur Hu, Bo-Ruei Huang, Juan Alvarez, Matthew Kim, Gabriel Margolis, Kavya Anbarasu, Masayoshi Tomizuka, Edward Adelson, Pulkit Agrawal
|
DEXOP: A Device for Robotic Transfer of Dexterous Human Manipulation
|
project page: https://dex-op.github.io
| null | null | null |
cs.RO cs.AI cs.CV cs.HC
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
We introduce perioperation, a paradigm for robotic data collection that sensorizes and records human manipulation while maximizing the transferability of the data to real robots. We implement this paradigm in DEXOP, a passive hand exoskeleton designed to maximize human ability to collect rich sensory (vision + tactile) data for diverse dexterous manipulation tasks in natural environments. DEXOP mechanically connects human fingers to robot fingers, providing users with direct contact feedback (via proprioception) and mirrors the human hand pose to the passive robot hand to maximize the transfer of demonstrated skills to the robot. The force feedback and pose mirroring make task demonstrations more natural for humans compared to teleoperation, increasing both speed and accuracy. We evaluate DEXOP across a range of dexterous, contact-rich tasks, demonstrating its ability to collect high-quality demonstration data at scale. Policies learned with DEXOP data significantly improve task performance per unit time of data collection compared to teleoperation, making DEXOP a powerful tool for advancing robot dexterity. Our project page is at https://dex-op.github.io.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 17:57:13 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 12:08:04 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Fang",
"Hao-Shu",
""
],
[
"Romero",
"Branden",
""
],
[
"Xie",
"Yichen",
""
],
[
"Hu",
"Arthur",
""
],
[
"Huang",
"Bo-Ruei",
""
],
[
"Alvarez",
"Juan",
""
],
[
"Kim",
"Matthew",
""
],
[
"Margolis",
"Gabriel",
""
],
[
"Anbarasu",
"Kavya",
""
],
[
"Tomizuka",
"Masayoshi",
""
],
[
"Adelson",
"Edward",
""
],
[
"Agrawal",
"Pulkit",
""
]
] |
2509.04482
|
Ravi Shankar
|
Ravi Shankar, Sheng Wong, Lin Li, Magdalena Bachmann, Alex Silverthorne, Beth Albert, Gabriel Davis Jones
|
Energy Landscapes Enable Reliable Abstention in Retrieval-Augmented Large Language Models for Healthcare
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Reliable abstention is critical for retrieval-augmented generation (RAG) systems, particularly in safety-critical domains such as women's health, where incorrect answers can lead to harm. We present an energy-based model (EBM) that learns a smooth energy landscape over a dense semantic corpus of 2.6M guideline-derived questions, enabling the system to decide when to generate or abstain. We benchmark the EBM against a calibrated softmax baseline and a k-nearest neighbour (kNN) density heuristic across both easy and hard abstention splits, where hard cases are semantically challenging near-distribution queries. The EBM achieves superior abstention performance abstention on semantically hard cases, reaching AUROC 0.961 versus 0.950 for softmax, while also reducing FPR@95 (0.235 vs 0.331). On easy negatives, performance is comparable across methods, but the EBM's advantage becomes most pronounced in safety-critical hard distributions. A comprehensive ablation with controlled negative sampling and fair data exposure shows that robustness stems primarily from the energy scoring head, while the inclusion or exclusion of specific negative types (hard, easy, mixed) sharpens decision boundaries but is not essential for generalisation to hard cases. These results demonstrate that energy-based abstention scoring offers a more reliable confidence signal than probability-based softmax confidence, providing a scalable and interpretable foundation for safe RAG systems.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 10:20:48 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 14:04:34 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Shankar",
"Ravi",
""
],
[
"Wong",
"Sheng",
""
],
[
"Li",
"Lin",
""
],
[
"Bachmann",
"Magdalena",
""
],
[
"Silverthorne",
"Alex",
""
],
[
"Albert",
"Beth",
""
],
[
"Jones",
"Gabriel Davis",
""
]
] |
2509.04484
|
Abdelrahman Sadallah
|
Abdelrahman Sadallah, Tim Baumg\"artner, Iryna Gurevych, Ted Briscoe
|
The Good, the Bad and the Constructive: Automatically Measuring Peer Review's Utility for Authors
|
EMNLP 2025 Main
| null | null | null |
cs.CL cs.AI cs.CY
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Providing constructive feedback to paper authors is a core component of peer review. With reviewers increasingly having less time to perform reviews, automated support systems are required to ensure high reviewing quality, thus making the feedback in reviews useful for authors. To this end, we identify four key aspects of review comments (individual points in weakness sections of reviews) that drive the utility for authors: Actionability, Grounding & Specificity, Verifiability, and Helpfulness. To enable evaluation and development of models assessing review comments, we introduce the RevUtil dataset. We collect 1,430 human-labeled review comments and scale our data with 10k synthetically labeled comments for training purposes. The synthetic data additionally contains rationales, i.e., explanations for the aspect score of a review comment. Employing the RevUtil dataset, we benchmark fine-tuned models for assessing review comments on these aspects and generating rationales. Our experiments demonstrate that these fine-tuned models achieve agreement levels with humans comparable to, and in some cases exceeding, those of powerful closed models like GPT-4o. Our analysis further reveals that machine-generated reviews generally underperform human reviews on our four aspects.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 14:19:07 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 06:37:53 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Sadallah",
"Abdelrahman",
""
],
[
"Baumgärtner",
"Tim",
""
],
[
"Gurevych",
"Iryna",
""
],
[
"Briscoe",
"Ted",
""
]
] |
2509.04537
|
Ryosuke Takata
|
Ryosuke Takata, Atsushi Masumori, Takashi Ikegami
|
Emergent Social Dynamics of LLM Agents in the El Farol Bar Problem
| null | null | null | null |
cs.MA cs.AI cs.CY
|
http://creativecommons.org/licenses/by/4.0/
|
We investigate the emergent social dynamics of Large Language Model (LLM) agents in a spatially extended El Farol Bar problem, observing how they autonomously navigate this classic social dilemma. As a result, the LLM agents generated a spontaneous motivation to go to the bar and changed their decision making by becoming a collective. We also observed that the LLM agents did not solve the problem completely, but rather behaved more like humans. These findings reveal a complex interplay between external incentives (prompt-specified constraints such as the 60% threshold) and internal incentives (culturally-encoded social preferences derived from pre-training), demonstrating that LLM agents naturally balance formal game-theoretic rationality with social motivations that characterize human behavior. These findings suggest that a new model of group decision making, which could not be handled in the previous game-theoretic problem setting, can be realized by LLM agents.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 08:09:42 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 11:56:01 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Takata",
"Ryosuke",
""
],
[
"Masumori",
"Atsushi",
""
],
[
"Ikegami",
"Takashi",
""
]
] |
2509.04650
|
Sharif Noor Zisad
|
Sharif Noor Zisad, N. M. Istiak Chowdhury and Ragib Hasan
|
Comparative Analysis of Transformer Models in Disaster Tweet Classification for Public Safety
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Twitter and other social media platforms have become vital sources of real time information during disasters and public safety emergencies. Automatically classifying disaster related tweets can help emergency services respond faster and more effectively. Traditional Machine Learning (ML) models such as Logistic Regression, Naive Bayes, and Support Vector Machines have been widely used for this task, but they often fail to understand the context or deeper meaning of words, especially when the language is informal, metaphorical, or ambiguous. We posit that, in this context, transformer based models can perform better than traditional ML models. In this paper, we evaluate the effectiveness of transformer based models, including BERT, DistilBERT, RoBERTa, and DeBERTa, for classifying disaster related tweets. These models are compared with traditional ML approaches to highlight the performance gap. Experimental results show that BERT achieved the highest accuracy (91%), significantly outperforming traditional models like Logistic Regression and Naive Bayes (both at 82%). The use of contextual embeddings and attention mechanisms allows transformer models to better understand subtle language in tweets, where traditional ML models fall short. This research demonstrates that transformer architectures are far more suitable for public safety applications, offering improved accuracy, deeper language understanding, and better generalization across real world social media text.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 20:22:33 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 16:28:10 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zisad",
"Sharif Noor",
""
],
[
"Chowdhury",
"N. M. Istiak",
""
],
[
"Hasan",
"Ragib",
""
]
] |
2509.04699
|
Wenhui Cui
|
Wenhui Cui, Christopher Sandino, Hadi Pouransari, Ran Liu, Juri Minxha, Ellen Zippi, Aman Verma, Anna Sedlackova, Erdrin Azemi, Behrooz Mahasseni
|
CPEP: Contrastive Pose-EMG Pre-training Enhances Gesture Generalization on EMG Signals
| null | null | null | null |
cs.LG eess.SP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Hand gesture classification using high-quality structured data such as videos, images, and hand skeletons is a well-explored problem in computer vision. Leveraging low-power, cost-effective biosignals, e.g. surface electromyography (sEMG), allows for continuous gesture prediction on wearables. In this paper, we demonstrate that learning representations from weak-modality data that are aligned with those from structured, high-quality data can improve representation quality and enables zero-shot classification. Specifically, we propose a Contrastive Pose-EMG Pre-training (CPEP) framework to align EMG and pose representations, where we learn an EMG encoder that produces high-quality and pose-informative representations. We assess the gesture classification performance of our model through linear probing and zero-shot setups. Our model outperforms emg2pose benchmark models by up to 21% on in-distribution gesture classification and 72% on unseen (out-of-distribution) gesture classification.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 23:12:17 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 06:09:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Cui",
"Wenhui",
""
],
[
"Sandino",
"Christopher",
""
],
[
"Pouransari",
"Hadi",
""
],
[
"Liu",
"Ran",
""
],
[
"Minxha",
"Juri",
""
],
[
"Zippi",
"Ellen",
""
],
[
"Verma",
"Aman",
""
],
[
"Sedlackova",
"Anna",
""
],
[
"Azemi",
"Erdrin",
""
],
[
"Mahasseni",
"Behrooz",
""
]
] |
2509.04700
|
Nisha Vinayaga-Sureshkanth
|
Mythili Menon, Nisha Vinayaga-Sureshkanth, Alec Schon, Kaitlyn Hemberger, Murtuza Jadliwala
|
Linguistic Hooks: Investigating The Role of Language Triggers in Phishing Emails Targeting African Refugees and Students
|
Mythili Menon and Nisha Vinayaga-Sureshkanth contributed equally to the work (co-first authors). Work accepted to Proceedings on Privacy Enhancing Technologies (PoPETs), Volume 2026, Issue 1
| null | null | null |
cs.CY
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Phishing and sophisticated email-based social engineering attacks disproportionately affect vulnerable populations, such as refugees and immigrant students. However, these groups remain understudied in cybersecurity research. This gap in understanding, coupled with their exclusion from broader security and privacy policies, increases their susceptibility to phishing and widens the digital security divide between marginalized and non-marginalized populations. To address this gap, we first conducted digital literacy workshops with newly resettled African refugee populations (n = 48) in the US to improve their understanding of how to safeguard sensitive and private information. Following the workshops, we conducted a real-world phishing deception study using carefully designed emails with linguistic cues for three participant groups: a subset of the African US-refugees recruited from the digital literacy workshops (n = 19), African immigrant students in the US (n = 142), and a control group of monolingual US-born students (n = 184). Our findings indicate that while digital literacy training for refugees improves awareness of safe cybersecurity practices, recently resettled African US-refugees still face significant challenges due to low digital literacy skills and limited English proficiency. This often leads them to ignore or fail to recognize phishing emails as phishing. Both African immigrant students and US-born students showed greater caution, though instances of data disclosure remained prevalent across groups. Our findings highlight, irrespective of literacy, the need to be trained to think critically about digital security. We conclude by discussing how the security and privacy community can better include marginalized populations in policy making and offer recommendations for designing equitable, inclusive cybersecurity initiatives.
|
[
{
"version": "v1",
"created": "Thu, 04 Sep 2025 23:13:04 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 15:43:47 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Menon",
"Mythili",
""
],
[
"Vinayaga-Sureshkanth",
"Nisha",
""
],
[
"Schon",
"Alec",
""
],
[
"Hemberger",
"Kaitlyn",
""
],
[
"Jadliwala",
"Murtuza",
""
]
] |
2509.04776
|
Jiahui Wang
|
J.-H. Wang, H. Xiong, J.-Z. Yang, H.-Y. Zhang, Y.-P. Song, L.-M. Duan
|
Transmon-assisted high-fidelity controlled-Z gates for integer fluxonium qubits
| null | null |
10.1103/qmds-z7gb
| null |
quant-ph
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Fluxoniums, as partially-protected superconducting qubits are promising to be employed to build high-performance large-scale quantum processor. The recently proposed ``integer fluxonium" operates at zero external flux bias, with a frequency of approximately 3 GHz. Single-qubit gate fidelity has been demonstrated to exceed $99.9\%$, while two-qubit gate schemes and scalable architectures remain underexplored. In this work, we investigate a fluxonium-transmon-fluxonium (FTF) coupling architecture using integer fluxoniums. We first confirm suppression of $ZZ$ interaction in the FTF system and then propose two high-fidelity controlled-$Z$ (CZ) gate schemes utilizing the coupler control: a flux-activated adiabatic gate scheme and a microwave-activated non-adiabatic gate scheme. Both schemes are capable of achieving low coherent error on the order of $1 \times 10^{-6}$ within gate durations of several tens of nanoseconds. Additionally, we discuss a hybrid circuit system in which an integer fluxonium is coupled to a conventional fluxonium through a transmon coupler. Our proposal provides insights for future implementations of large-scale quantum circuits based on integer fluxonium devices.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 03:12:54 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wang",
"J. -H.",
""
],
[
"Xiong",
"H.",
""
],
[
"Yang",
"J. -Z.",
""
],
[
"Zhang",
"H. -Y.",
""
],
[
"Song",
"Y. -P.",
""
],
[
"Duan",
"L. -M.",
""
]
] |
2509.04783
|
Tong Lin
|
Ziyang Xiong, Tong Lin, Liu Li, Hao Deng, Shihua Chen, Junpeng Lu, and Zhenhua Ni
|
Million-Q Dual-Polarization Micro-Fabry-Perot Resonators in Silicon Nitride Photonic Integrated Circuits
|
9 pages,6 figures
|
Optics & Laser Technology,Volume 192, Part D,2025,113854,ISSN 0030-3992
|
10.1016/j.optlastec.2025.113854
| null |
physics.optics
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Miniaturized Fabry-Perot standing-wave resonators and whispering-gallery travelling wave resonators constitute foundational building blocks for photonic integrated circuits. While both architectures offer transformative potential through high quality factors and dual-polarization operation, integrated Fabry-Perot resonators face significant challenges in simultaneously achieving ultra-high Q-factors and broadband thermal tunability for fundamental transverse magnetic (TM0) and transverse electric (TE0) modes within a compact footprint-primarily due to polarization-dependent losses in conventional chip-scale reflectors. Here, we overcome this limitation by demonstrating an integrated silicon nitride dual-polarization micro-Fabry-Perot resonator with polarization-insensitive Sagnac loop reflectors and multimode waveguides to effectively suppress losses and enable high-performances for both fundamental transverse magnetic (TM0) and transverse electric (TE0) modes. The device achieves record loaded quality factors of 2.38*106 (TM0) and 3.48*105 (TE0) respectively and intrinsic quality factors will be even higher. Moreover, both two modes are tuned over the whole free spectral range of around 0.111 nm (TM0) and 0.112 nm (TE0) with the thermal tuning efficiencies of approximately 1.04 pm/mW (TM0) and 1.24 pm/mW (TE0). These advances establish a new benchmark for compact, high-performance dual-polarization resonators in optical sensors, nonlinear and integrated quantum photonics.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 03:36:37 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 15:03:05 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Xiong",
"Ziyang",
""
],
[
"Lin",
"Tong",
""
],
[
"Li",
"Liu",
""
],
[
"Deng",
"Hao",
""
],
[
"Chen",
"Shihua",
""
],
[
"Lu",
"Junpeng",
""
],
[
"Ni",
"Zhenhua",
""
]
] |
2509.04809
|
Haechang Kim
|
Haechang Kim, Hao Chen, Can Li, Jong Min Lee
|
TalkToAgent: A Human-centric Explanation of Reinforcement Learning Agents with Large Language Models
|
31 pages total
| null | null | null |
cs.AI cs.HC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Explainable Reinforcement Learning (XRL) has emerged as a promising approach in improving the transparency of Reinforcement Learning (RL) agents. However, there remains a gap between complex RL policies and domain experts, due to the limited comprehensibility of XRL results and isolated coverage of current XRL approaches that leave users uncertain about which tools to employ. To address these challenges, we introduce TalkToAgent, a multi-agent Large Language Models (LLM) framework that delivers interactive, natural language explanations for RL policies. The architecture with five specialized LLM agents (Coordinator, Explainer, Coder, Evaluator, and Debugger) enables TalkToAgent to automatically map user queries to relevant XRL tools and clarify an agent's actions in terms of either key state variables, expected outcomes, or counterfactual explanations. Moreover, our approach extends previous counterfactual explanations by deriving alternative scenarios from qualitative behavioral descriptions, or even new rule-based policies. We validated TalkToAgent on quadruple-tank process control problem, a well-known nonlinear control benchmark. Results demonstrated that TalkToAgent successfully mapped user queries into XRL tasks with high accuracy, and coder-debugger interactions minimized failures in counterfactual generation. Furthermore, qualitative evaluation confirmed that TalkToAgent effectively interpreted agent's actions and contextualized their meaning within the problem domain.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 05:09:09 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 00:52:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kim",
"Haechang",
""
],
[
"Chen",
"Hao",
""
],
[
"Li",
"Can",
""
],
[
"Lee",
"Jong Min",
""
]
] |
2509.04817
|
Petar Mlinari\'c
|
Petar Mlinari\'c, Serkan Gugercin, Zoran Tomljanovi\'c
|
Optimal Damping for the 1D Wave Equation Using a Single Damper
|
18 pages, 12 figures
| null | null | null |
math.OC cs.SY eess.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vibrational structures are susceptible to catastrophic failures or structural damages when external forces induce resonances or repeated unwanted oscillations. One common mitigation strategy is to use dampers to suppress these disturbances. This leads to the problem of finding optimal damper viscosities and positions for a given vibrational structure. Although extensive research exists for the case of finite-dimensional systems, optimizing damper positions remains challenging due to its discrete nature. To overcome this, we introduce a novel model for the damped wave equation (at the PDE level) with a damper of viscosity $\mathfrak{g}$ at position $\mathfrak{p}$ and develop a system-theoretic input/output-based analysis in the frequency domain. In this system-theoretic formulation, while we consider average displacement as the output, for input (forcing), we analyze two separate cases, namely, the uniform and boundary forcing. For both cases, explicit formulas are derived for the corresponding transfer functions, parametrized by $\mathfrak{p}$ and $\mathfrak{g}$. This explicit parametrization by $\mathfrak{p}$ and $\mathfrak{g}$ facilitates analyzing the optimal damping problem (at the PDE level) using norms such as the $\mathcal{H}_2$ and $\mathcal{H}_\infty$ norms. We also examine limiting cases, such as when the viscosity is very large or when no external damping is present. To illustrate our approach, we present numerical examples, compare different optimization criteria, and discuss the impact of damping parameters on the damped wave equation.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 05:37:54 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 14:38:50 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Mlinarić",
"Petar",
""
],
[
"Gugercin",
"Serkan",
""
],
[
"Tomljanović",
"Zoran",
""
]
] |
2509.04819
|
Shuhan Ding
|
Shuhan Ding, Jingjing Fu, Yu Gu, Naiteek Sangani, Mu Wei, Paul Vozila, Nan Liu, Jiang Bian, Hoifung Poon
|
AURAD: Anatomy-Pathology Unified Radiology Synthesis with Progressive Representations
| null | null | null | null |
eess.IV cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Medical image synthesis has become an essential strategy for augmenting datasets and improving model generalization in data-scarce clinical settings. However, fine-grained and controllable synthesis remains difficult due to limited high-quality annotations and domain shifts across datasets. Existing methods, often designed for natural images or well-defined tumors, struggle to generalize to chest radiographs, where disease patterns are morphologically diverse and tightly intertwined with anatomical structures. To address these challenges, we propose AURAD, a controllable radiology synthesis framework that jointly generates high-fidelity chest X-rays and pseudo semantic masks. Unlike prior approaches that rely on randomly sampled masks-limiting diversity, controllability, and clinical relevance-our method learns to generate masks that capture multi-pathology coexistence and anatomical-pathological consistency. It follows a progressive pipeline: pseudo masks are first generated from clinical prompts conditioned on anatomical structures, and then used to guide image synthesis. We also leverage pretrained expert medical models to filter outputs and ensure clinical plausibility. Beyond visual realism, the synthesized masks also serve as labels for downstream tasks such as detection and segmentation, bridging the gap between generative modeling and real-world clinical applications. Extensive experiments and blinded radiologist evaluations demonstrate the effectiveness and generalizability of our method across tasks and datasets. In particular, 78% of our synthesized images are classified as authentic by board-certified radiologists, and over 40% of predicted segmentation overlays are rated as clinically useful. All code, pre-trained models, and the synthesized dataset will be released upon publication.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 05:40:55 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 16:46:10 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ding",
"Shuhan",
""
],
[
"Fu",
"Jingjing",
""
],
[
"Gu",
"Yu",
""
],
[
"Sangani",
"Naiteek",
""
],
[
"Wei",
"Mu",
""
],
[
"Vozila",
"Paul",
""
],
[
"Liu",
"Nan",
""
],
[
"Bian",
"Jiang",
""
],
[
"Poon",
"Hoifung",
""
]
] |
2509.04895
|
Maryam Adelipour
|
Maryam Adelipour, Gustavo Carneiro, Jeongkwon Kim
|
Evaluating Multiple Instance Learning Strategies for Automated Sebocyte Droplet Counting
|
8 pages, 1 figure, 2 tables
| null | null | null |
cs.CV cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Sebocytes are lipid-secreting cells whose differentiation is marked by the accumulation of intracellular lipid droplets, making their quantification a key readout in sebocyte biology. Manual counting is labor-intensive and subjective, motivating automated solutions. Here, we introduce a simple attention-based multiple instance learning (MIL) framework for sebocyte image analysis. Nile Red-stained sebocyte images were annotated into 14 classes according to droplet counts, expanded via data augmentation to about 50,000 cells. Two models were benchmarked: a baseline multi-layer perceptron (MLP) trained on aggregated patch-level counts, and an attention-based MIL model leveraging ResNet-50 features with instance weighting. Experiments using five-fold cross-validation showed that the baseline MLP achieved more stable performance (mean MAE = 5.6) compared with the attention-based MIL, which was less consistent (mean MAE = 10.7) but occasionally superior in specific folds. These findings indicate that simple bag-level aggregation provides a robust baseline for slide-level droplet counting, while attention-based MIL requires task-aligned pooling and regularization to fully realize its potential in sebocyte image analysis.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 08:15:57 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Adelipour",
"Maryam",
""
],
[
"Carneiro",
"Gustavo",
""
],
[
"Kim",
"Jeongkwon",
""
]
] |
2509.04899
|
Hiroshi Watanabe
|
Mutsumi Kobayashi and Hiroshi Watanabe
|
Learning and composing of classical music using restricted Boltzmann machines
|
19 pages, 10 figures, Figures are updated
| null | null | null |
cs.SD cs.LG eess.AS
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recently, software has been developed that uses machine learning to mimic the style of a particular composer, such as J. S. Bach. However, since such software often adopts machine learning models with complex structures, it is difficult to analyze how the software understands the characteristics of the composer's music. In this study, we adopted J. S. Bach's music for training of a restricted Boltzmann machine (RBM). Since the structure of RBMs is simple, it allows us to investigate the internal states after learning. We found that the learned RBM is able to compose music.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 08:18:14 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 07:31:38 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Kobayashi",
"Mutsumi",
""
],
[
"Watanabe",
"Hiroshi",
""
]
] |
2509.04904
|
Gianmarco Caruso
|
G. Caruso, W. F. Rosenberger, P. Mozgunov, N. Flournoy
|
An information metric for comparing and assessing informative interim decisions in sequential clinical trials
| null | null | null | null |
stat.ME cs.IT math.IT stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Group sequential designs enable interim analyses and potential early stopping for efficacy or futility. While these adaptations improve trial efficiency and ethical considerations, they also introduce bias into the adapted analyses. We demonstrate how failing to account for informative interim decisions in the analysis can substantially affect posterior estimates of the treatment effect, often resulting in overly optimistic credible intervals aligned with the stopping decision. Drawing on information theory, we use the Kullback-Leibler divergence to quantify this distortion and highlight its use for post-hoc evaluation of informative interim decisions, with a focus on end-of-study inference. Unlike pointwise comparisons, this measure provides an integrated summary of this distortion on the whole parameter space. By comparing alternative decision boundaries and prior specifications, we illustrate how this measure can improve the understanding of trial results and inform the planning of future adaptive studies. We also introduce an expected version of this metric to support clinicians in choosing decision boundaries. This guidance complements traditional strategies based on type-I error rate control by offering insights into the distortion introduced to the treatment effect at each interim phase. The use of this pre-experimental measure is finally illustrated in a group sequential trial for evaluating a treatment for central nervous system disorders.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 08:23:14 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 13:10:02 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Caruso",
"G.",
""
],
[
"Rosenberger",
"W. F.",
""
],
[
"Mozgunov",
"P.",
""
],
[
"Flournoy",
"N.",
""
]
] |
2509.04947
|
Anran Jin
|
Anran Jin and Ling-Yan Hung
|
Note on searching for critical lattice models as entropy critical points from strange correlator
|
10 pages, 7 figures. Comments are welcome
| null | null | null |
cond-mat.str-el hep-th quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
An entropy function is proposed in [Phys. Rev. Lett. 131, 251602] as a way to detect criticality even when the system size is small. In this note we apply this strategy in the search for criticality of lattice transfer matrices constructed based on the topological holographic principle. We find that the combination of strategy is indeed a cost-effective and efficient way of identifying critical boundary conditions, estimating central charges and moreover, plotting entire phase diagrams in a multi-dimensional phase space.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 09:14:50 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 14:42:24 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Jin",
"Anran",
""
],
[
"Hung",
"Ling-Yan",
""
]
] |
2509.04955
|
Guolong Zhong
|
Guolong Zhong, Yi Fan, Zhenyu Li
|
Scalable parallel simulation of quantum circuits on CPU and GPU systems
| null | null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
Quantum computing enables parallelism through superposition and entanglement and offers advantages over classical computing architectures. However, due to the limitations of current quantum hardware in the noisy intermediate-scale quantum (NISQ) era, classical simulation remains a critical tool for developing quantum algorithms. In this research, we present a comprehensive parallelization solution for the Q$^2$Chemistry software package, delivering significant performance improvements for the full-amplitude simulator on both CPU and GPU platforms. By incorporating batch-buffered overlap processing, dependency-aware gate contraction and staggered multi-gate parallelism, our optimizations significantly enhance the simulation speed compared to unoptimized baselines, demonstrating the effectiveness of hybrid-level parallelism in HPC systems. Benchmark results show that Q$^2$Chemistry consistently outperforms current state-of-the-art open-source simulators across various circuit types. These benchmarks highlight the capability of Q$^2$Chemistry to effectively handle large-scale quantum simulations with high efficiency and high portability.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 09:20:11 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 14:48:09 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhong",
"Guolong",
""
],
[
"Fan",
"Yi",
""
],
[
"Li",
"Zhenyu",
""
]
] |
2509.05007
|
Jie Chen
|
Jie Chen, Jinhao Jiang, Yingqian Min, Zican Dong, Shijie Wang, Wayne Xin Zhao, Ji-Rong Wen
|
Sticker-TTS: Learn to Utilize Historical Experience with a Sticker-driven Test-Time Scaling Framework
|
11 pages, 1 figures, 5 tables
| null | null | null |
cs.AI cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large reasoning models (LRMs) have exhibited strong performance on complex reasoning tasks, with further gains achievable through increased computational budgets at inference. However, current test-time scaling methods predominantly rely on redundant sampling, ignoring the historical experience utilization, thereby limiting computational efficiency. To overcome this limitation, we propose Sticker-TTS, a novel test-time scaling framework that coordinates three collaborative LRMs to iteratively explore and refine solutions guided by historical attempts. At the core of our framework are distilled key conditions-termed stickers-which drive the extraction, refinement, and reuse of critical information across multiple rounds of reasoning. To further enhance the efficiency and performance of our framework, we introduce a two-stage optimization strategy that combines imitation learning with self-improvement, enabling progressive refinement. Extensive evaluations on three challenging mathematical reasoning benchmarks, including AIME-24, AIME-25, and OlymMATH, demonstrate that Sticker-TTS consistently surpasses strong baselines, including self-consistency and advanced reinforcement learning approaches, under comparable inference budgets. These results highlight the effectiveness of sticker-guided historical experience utilization. Our code and data are available at https://github.com/RUCAIBox/Sticker-TTS.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 11:14:11 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 03:26:03 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chen",
"Jie",
""
],
[
"Jiang",
"Jinhao",
""
],
[
"Min",
"Yingqian",
""
],
[
"Dong",
"Zican",
""
],
[
"Wang",
"Shijie",
""
],
[
"Zhao",
"Wayne Xin",
""
],
[
"Wen",
"Ji-Rong",
""
]
] |
2509.05037
|
Noorul Wahab
|
Noorul Wahab, Ethar Alzaid, Jiaqi Lv, Adam Shephard, Shan E Ahmed Raza
|
ModalSurv: A Multimodal Deep Survival Framework for Prostrate and Bladder Cancer
|
6 pages, 1 figure, 2 tables
| null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Accurate prediction of time-to-event outcomes is a central challenge in oncology, with significant implications for treatment planning and patient management. In this work, we present ModaliSurv, a multimodal deep survival model utilising DeepHit with a projection layer and inter-modality cross-attention, which integrates heterogeneous patient data, including clinical, MRI, RNA-seq and whole-slide pathology features. The model is designed to capture complementary prognostic signals across modalities and estimate individualised time-to-biochemical recurrence in prostate cancer and time-to-cancer recurrence in bladder cancer. Our approach was evaluated in the context of the CHIMERA Grand Challenge, across two of the three provided tasks. For Task 1 (prostate cancer bio-chemical recurrence prediction), the proposed framework achieved a concordance index (C-index) of 0.843 on 5-folds cross-validation and 0.818 on CHIMERA development set, demonstrating robust discriminatory ability. For Task 3 (bladder cancer recurrence prediction), the model obtained a C-index of 0.662 on 5-folds cross-validation and 0.457 on development set, highlighting its adaptability and potential for clinical translation. These results suggest that leveraging multimodal integration with deep survival learning provides a promising pathway toward personalised risk stratification in prostate and bladder cancer. Beyond the challenge setting, our framework is broadly applicable to survival prediction tasks involving heterogeneous biomedical data.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 11:52:53 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 12:06:11 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wahab",
"Noorul",
""
],
[
"Alzaid",
"Ethar",
""
],
[
"Lv",
"Jiaqi",
""
],
[
"Shephard",
"Adam",
""
],
[
"Raza",
"Shan E Ahmed",
""
]
] |
2509.05050
|
Andrey Voronov
|
Andrey A. Voronov, Khrystyna O. Levchenko, Roman Verba, Krist\'yna Dav\'idkov\'a, Carsten Dubs, Michal Urb\'anek, Qi Wang, Dieter Suess, Claas Abert, and Andrii V. Chumak
|
Exchange spin-wave propagation in Ga:YIG nanowaveguides
| null | null | null | null |
cond-mat.other
|
http://creativecommons.org/licenses/by/4.0/
|
Spin-wave-based computing has emerged as a promising approach to overcome the fundamental limitations of CMOS technologies. However, the increasing demand for device miniaturization down to a 100 nm scale presents significant challenges for long-distance spin-wave transport. Gallium-substituted yttrium iron garnet (Ga:YIG) offers a potential solution to these challenges due to its unique magnetic properties. The reduced saturation magnetization in Ga:YIG enables efficient excitation of exchange-dominated spin waves, which exhibit enhanced transport characteristics compared to dipolar-dominated modes in conventional materials. Here, we present the first comprehensive study combining experimental, analytical, and numerical investigations of spin-wave propagation in Ga:YIG waveguides down to 145 nm width and 73 nm thickness. Using micro-focused Brillouin light scattering spectroscopy, TetraX simulations, and analytical dispersion calculations, we demonstrate that Ga:YIG waveguides support spin waves with significantly higher group velocities up to 600 m/s. This value remains constant for structures with different widths, leading to longer spin-wave propagation lengths in nanowaveguides compared to non-substituted YIG. These results reveal that gallium substitution provides access to faster and longer-lived spin waves, opening new possibilities for implementing this material in nanoscale magnonic devices.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 12:31:23 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 10:16:19 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Voronov",
"Andrey A.",
""
],
[
"Levchenko",
"Khrystyna O.",
""
],
[
"Verba",
"Roman",
""
],
[
"Davídková",
"Kristýna",
""
],
[
"Dubs",
"Carsten",
""
],
[
"Urbánek",
"Michal",
""
],
[
"Wang",
"Qi",
""
],
[
"Suess",
"Dieter",
""
],
[
"Abert",
"Claas",
""
],
[
"Chumak",
"Andrii V.",
""
]
] |
2509.05052
|
Marek Li\v{s}ka
|
Ana Alonso-Serrano, Marek Li\v{s}ka and Micha{\l} Piotrak
|
Semi-classical spacetime thermodynamics
|
36 pages
| null | null |
DIAS-STP-25-19
|
hep-th gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We derive the semi-classical gravitational dynamics from thermodynamics of local stretched light cones in 2-dimensional dilaton gravity, explicitly treating the backreaction of quantum matter through the conformal anomaly's effect on the generalized entropy. We also sketch the extension of this analysis to the conformal anomaly in 4-dimensional semi-classical gravity. In direct connection to this problem, we also tackle the appropriate definition of Wald entropy in thermodynamic derivation of equations of motion for classical scalar-tensor theories. For the class of Brans-Dicke theories, including 2-dimensional dilaton gravity, we show that the equations of motion follow from the dynamical Wald entropy associated with local causal horizons.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 12:31:59 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Alonso-Serrano",
"Ana",
""
],
[
"Liška",
"Marek",
""
],
[
"Piotrak",
"Michał",
""
]
] |
2509.05139
|
Paolo Pareti Dr.
|
Jaime Osvaldo Salas, Paolo Pareti, Semih Yumu\c{s}ak, Soulmaz Gheisari, Luis-Daniel Ib\'a\~nez, George Konstantinidis
|
Evaluation and Comparison Semantics for ODRL
|
Accepted as a full paper at the 14th International Joint Conference on Knowledge Graphs (IJCKG 2025). This is the submitted manuscript, the accepted manuscript will be published by Springer Nature
| null | null | null |
cs.AI cs.LO
|
http://creativecommons.org/licenses/by/4.0/
|
We consider the problem of evaluating, and comparing computational policies in the Open Digital Rights Language (ODRL), which has become the de facto standard for governing the access and usage of digital resources. Although preliminary progress has been made on the formal specification of the language's features, a comprehensive formal semantics of ODRL is still missing. In this paper, we provide a simple and intuitive formal semantics for ODRL that is based on query answering. Our semantics refines previous formalisations, and is aligned with the latest published specification of the language (2.2). Building on our evaluation semantics, and motivated by data sharing scenarios, we also define and study the problem of comparing two policies, detecting equivalent, more restrictive or more permissive policies.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 14:30:41 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Salas",
"Jaime Osvaldo",
""
],
[
"Pareti",
"Paolo",
""
],
[
"Yumuşak",
"Semih",
""
],
[
"Gheisari",
"Soulmaz",
""
],
[
"Ibáñez",
"Luis-Daniel",
""
],
[
"Konstantinidis",
"George",
""
]
] |
2509.05149
|
Hung Ho Huy
|
Huy Hung Ho, Nhan Le Thanh, and Nam Nguyen Hong
|
Odoo-based Subcontract Inter-site Access Control Mechanism for Construction Projects
|
7 pages, 2 figures, The 9th OISP Science and Technology Symposium
| null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In the era of Construction 4.0, the industry is embracing a new paradigm of labor elasticity, driven by smart and flexible outsourcing and subcontracting strategies. The increased reliance on specialized subcontractors enables companies to scale labor dynamically based on project demands. This adaptable workforce model presents challenges in managing hierarchical integration and coordinating inter-site collaboration. Our design introduces a subsystem integrated into the Odoo ERP framework, employing a modular architecture to streamline labor management, task tracking, and approval workflows. The system adopts a three-pronged approach to ensure synchronized data exchange between general contractors and subcontractors, while maintaining both security and operational independence. The system features hybrid access control, third-party integration for cross-domain communication, and role-based mapping algorithm across sites. The system supports varying degrees of customization through a unified and consolidated attribute mapping center. This center leverages a tree-like index structure and Lagrange interpolation method to enhance the efficiency of role mapping. Demonstrations highlight practical application in outsourcing, integration, and scalability scenarios, confirming the system's robustness under high user volumes and in offline conditions. Experimental results further show improvements in database performance and workflow adaptability to support a scalable, enterprise-level solution that aligns with the evolving demands of smart construction management.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 14:38:19 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 14:44:05 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Ho",
"Huy Hung",
""
],
[
"Thanh",
"Nhan Le",
""
],
[
"Hong",
"Nam Nguyen",
""
]
] |
2509.05155
|
Yi Zhang
|
Hongjie Dong and Yi Ru-Ya Zhang
|
Serrin's overdetermined theorem within Lipschitz domains
|
11 pages, the abstract was corrected
| null | null | null |
math.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Let $\Omega\subset\mathbb R^n$ be a Lipschitz domain, $K$ be a (bounded) ellipsoid centered at the origin and $H$ be the associated Wulff potential. We prove that, $\Omega$ satisfies the following Serrin-type overdetermined system
$$u \in W^{1,2}(\mathbb R^n), \quad u=0\ \text{ a.e. in }\mathbb R^n\setminus \Omega,\quad \Delta_H u=\mathbf{c}\mathscr{H}^{n-1}|_{\partial^*\Omega} - \mathbf{1}_{\Omega}\,dx,$$
in the weak sense if and only if $\Omega$ is homothetic to $K$. Here $\Delta_H$ denotes the anisotropic Laplacian associated to $H$, and $\mathscr H^{n-1}$ denotes the $(n-1)$-dimensional Hausdorff measure. Our approach offers an alternative proof to [11] in the case of Lipschitz domains, introducing a novel viewpoint to settle [13, Question 7.1].
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 14:48:36 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 15:40:14 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Dong",
"Hongjie",
""
],
[
"Zhang",
"Yi Ru-Ya",
""
]
] |
2509.05166
|
Sujit Kumar Sikder
|
Sujit Kumar Sikder, Jyotirmaya Ijaradar, Hao Li, Hichem Omrani
|
Transition of car-based human-mobility in the pandemic era: Data insight from a cross-border region in Europe
| null | null | null | null |
cs.HC cs.CY
|
http://creativecommons.org/licenses/by/4.0/
|
Many transport authorities are collecting and publishing almost real-time road traffic data to meet the growing trend of massive open data, a vital resource for foresight decision support systems considering deep data insights. We explored the spatio-temporal transitions in the cross-country road traffic volumes in the context of modelling behavioural transitions in car-based human mobility. This study reports on individual car-based daily travel behaviour detected, before (2018) and during the COVID pandemic (2020), between Germany and neighbouring countries. In the case of Luxembourg, the Bridges and Roads Authority has installed a large digital traffic observatory infrastructure through the adoption of sensor-based IoT technologies, like other European member states. Since 2016, they have provided high-performance data processing and published open data on the country's road traffic. The dataset contains an hourly traffic count for different vehicle types, daily for representative observation points, followed by a major road network. The original dataset contains significant missing entries, so comprehensive data harmonization was performed. We observed the decrease in traffic volumes during pandemic factors (e.g. lockdowns and remote work) period by following global trend of reduced personal mobility. The understanding the dynamic adaptive travel behaviours provide a potential opportunity to generate the actionable insight including temporal and spatial implications. This study demonstrates that the national open traffic data products can have adoption potential to address cross-border insights. In relevance to the net-zero carbon transition, further study should shed light on the interpolation and downscaling approaches at the comprehensive road-network level for identifying pollution hot spots, causal link to functional landuse patterns and calculation of spatial influence area.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 15:03:22 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 11:09:04 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Sikder",
"Sujit Kumar",
""
],
[
"Ijaradar",
"Jyotirmaya",
""
],
[
"Li",
"Hao",
""
],
[
"Omrani",
"Hichem",
""
]
] |
2509.05186
|
Benjamin Zhang
|
Benjamin J. Zhang, Siting Liu, Stanley J. Osher, Markos A. Katsoulakis
|
Probabilistic operator learning: generative modeling and uncertainty quantification for foundation models of differential equations
|
First two authors contributed equally
| null | null | null |
stat.ML cs.LG cs.NA math.NA
|
http://creativecommons.org/licenses/by/4.0/
|
In-context operator networks (ICON) are a class of operator learning methods based on the novel architectures of foundation models. Trained on a diverse set of datasets of initial and boundary conditions paired with corresponding solutions to ordinary and partial differential equations (ODEs and PDEs), ICON learns to map example condition-solution pairs of a given differential equation to an approximation of its solution operator. Here, we present a probabilistic framework that reveals ICON as implicitly performing Bayesian inference, where it computes the mean of the posterior predictive distribution over solution operators conditioned on the provided context, i.e., example condition-solution pairs. The formalism of random differential equations provides the probabilistic framework for describing the tasks ICON accomplishes while also providing a basis for understanding other multi-operator learning methods. This probabilistic perspective provides a basis for extending ICON to \emph{generative} settings, where one can sample from the posterior predictive distribution of solution operators. The generative formulation of ICON (GenICON) captures the underlying uncertainty in the solution operator, which enables principled uncertainty quantification in the solution predictions in operator learning.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 15:35:04 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 17:28:39 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhang",
"Benjamin J.",
""
],
[
"Liu",
"Siting",
""
],
[
"Osher",
"Stanley J.",
""
],
[
"Katsoulakis",
"Markos A.",
""
]
] |
2509.05192
|
Simon Lachnit
|
Simon Lachnit and Ghassan Karame
|
On Hyperparameters and Backdoor-Resistance in Horizontal Federated Learning
|
To appear in the Proceedings of the ACM Conference on Computer and Communications Security (CCS) 2025
| null | null | null |
cs.CR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Horizontal Federated Learning (HFL) is particularly vulnerable to backdoor attacks as adversaries can easily manipulate both the training data and processes to execute sophisticated attacks. In this work, we study the impact of training hyperparameters on the effectiveness of backdoor attacks and defenses in HFL. More specifically, we show both analytically and by means of measurements that the choice of hyperparameters by benign clients does not only influence model accuracy but also significantly impacts backdoor attack success. This stands in sharp contrast with the multitude of contributions in the area of HFL security, which often rely on custom ad-hoc hyperparameter choices for benign clients$\unicode{x2013}$leading to more pronounced backdoor attack strength and diminished impact of defenses. Our results indicate that properly tuning benign clients' hyperparameters$\unicode{x2013}$such as learning rate, batch size, and number of local epochs$\unicode{x2013}$can significantly curb the effectiveness of backdoor attacks, regardless of the malicious clients' settings. We support this claim with an extensive robustness evaluation of state-of-the-art attack-defense combinations, showing that carefully chosen hyperparameters yield across-the-board improvements in robustness without sacrificing main task accuracy. For example, we show that the 50%-lifespan of the strong A3FL attack can be reduced by 98.6%, respectively$\unicode{x2013}$all without using any defense and while incurring only a 2.9 percentage points drop in clean task accuracy.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 15:46:03 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 13:52:25 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Lachnit",
"Simon",
""
],
[
"Karame",
"Ghassan",
""
]
] |
2509.05218
|
Hongyu Shan Shhy
|
Chang Dai, Hongyu Shan, Mingyang Song, Di Liang
|
HoPE: Hyperbolic Rotary Positional Encoding for Stable Long-Range Dependency Modeling in Large Language Models
| null | null | null | null |
cs.CL cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Positional encoding mechanisms enable Transformers to model sequential structure and long-range dependencies in text. While absolute positional encodings struggle with extrapolation to longer sequences due to fixed positional representations, and relative approaches like Alibi exhibit performance degradation on extremely long contexts, the widely-used Rotary Positional Encoding (RoPE) introduces oscillatory attention patterns that hinder stable long-distance dependency modelling. We address these limitations through a geometric reformulation of positional encoding. Drawing inspiration from Lorentz transformations in hyperbolic geometry, we propose Hyperbolic Rotary Positional Encoding (HoPE), which leverages hyperbolic functions to implement Lorentz rotations on token representations. Theoretical analysis demonstrates that RoPE is a special case of our generalized formulation. HoPE fundamentally resolves RoPE's slation issues by enforcing monotonic decay of attention weights with increasing token distances. Extensive experimental results, including perplexity evaluations under several extended sequence benchmarks, show that HoPE consistently exceeds existing positional encoding methods. These findings underscore HoPE's enhanced capacity for representing and generalizing long-range dependencies. Data and code will be available.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 16:20:48 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 03:13:38 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Dai",
"Chang",
""
],
[
"Shan",
"Hongyu",
""
],
[
"Song",
"Mingyang",
""
],
[
"Liang",
"Di",
""
]
] |
2509.05231
|
Mathilde Andr\'e
|
Mathilde Andr\'e, F\'elix Foutel-Rodier and Emmanuel Schertzer
|
Moments of density-dependent branching processes and their genealogy
|
28 pages, 1 figure
| null | null | null |
math.PR
|
http://creativecommons.org/licenses/by/4.0/
|
A density-dependent branching process is a particle system in which individuals reproduce independently, but in a way that depends on the current population size. This feature can model a wide range of ecological interactions at the cost of breaking the branching property. We propose a general approach for studying the genealogy of these models based on moments. Building on a recent work of Bansaye, we show how to compute recursively these moments in a similar spirit to the many-to-few formula in the theory of branching processes. These formulas enable one to deduce the convergence of the genealogy by studying the population density, for which stochastic calculus techniques are available. As a first application of these ideas, we consider a density-dependent branching process started close to a stable equilibrium of the ecological dynamics. We show that, under a finite second moment assumption, its genealogy converges to Kingman's coalescent when the carrying capacity of the population goes to infinity.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 16:47:53 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 11:27:31 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"André",
"Mathilde",
""
],
[
"Foutel-Rodier",
"Félix",
""
],
[
"Schertzer",
"Emmanuel",
""
]
] |
2509.05263
|
Shuang Qiu
|
Yinglin Duan, Zhengxia Zou, Tongwei Gu, Wei Jia, Zhan Zhao, Luyi Xu, Xinzhu Liu, Yenan Lin, Hao Jiang, Kang Chen, Shuang Qiu
|
LatticeWorld: A Multimodal Large Language Model-Empowered Framework for Interactive Complex World Generation
| null | null | null | null |
cs.AI cs.CV cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent research has been increasingly focusing on developing 3D world models that simulate complex real-world scenarios. World models have found broad applications across various domains, including embodied AI, autonomous driving, entertainment, etc. A more realistic simulation with accurate physics will effectively narrow the sim-to-real gap and allow us to gather rich information about the real world conveniently. While traditional manual modeling has enabled the creation of virtual 3D scenes, modern approaches have leveraged advanced machine learning algorithms for 3D world generation, with most recent advances focusing on generative methods that can create virtual worlds based on user instructions. This work explores such a research direction by proposing LatticeWorld, a simple yet effective 3D world generation framework that streamlines the industrial production pipeline of 3D environments. LatticeWorld leverages lightweight LLMs (LLaMA-2-7B) alongside the industry-grade rendering engine (e.g., Unreal Engine 5) to generate a dynamic environment. Our proposed framework accepts textual descriptions and visual instructions as multimodal inputs and creates large-scale 3D interactive worlds with dynamic agents, featuring competitive multi-agent interaction, high-fidelity physics simulation, and real-time rendering. We conduct comprehensive experiments to evaluate LatticeWorld, showing that it achieves superior accuracy in scene layout generation and visual fidelity. Moreover, LatticeWorld achieves over a $90\times$ increase in industrial production efficiency while maintaining high creative quality compared with traditional manual production methods. Our demo video is available at https://youtu.be/8VWZXpERR18
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 17:22:33 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 17:05:47 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Duan",
"Yinglin",
""
],
[
"Zou",
"Zhengxia",
""
],
[
"Gu",
"Tongwei",
""
],
[
"Jia",
"Wei",
""
],
[
"Zhao",
"Zhan",
""
],
[
"Xu",
"Luyi",
""
],
[
"Liu",
"Xinzhu",
""
],
[
"Lin",
"Yenan",
""
],
[
"Jiang",
"Hao",
""
],
[
"Chen",
"Kang",
""
],
[
"Qiu",
"Shuang",
""
]
] |
2509.05289
|
Martina Boschi
|
Martina Boschi and J\"urgen Lerner and Ernst C. Wit
|
Beyond Linearity and Time-homogeneity: Relational Hyper Event Models with Time-Varying Non-Linear Effects
| null | null | null | null |
stat.ME cs.LG stat.AP
|
http://creativecommons.org/licenses/by/4.0/
|
Recent technological advances have made it easier to collect large and complex networks of time-stamped relational events connecting two or more entities. Relational hyper-event models (RHEMs) aim to explain the dynamics of these events by modeling the event rate as a function of statistics based on past history and external information.
However, despite the complexity of the data, most current RHEM approaches still rely on a linearity assumption to model this relationship. In this work, we address this limitation by introducing a more flexible model that allows the effects of statistics to vary non-linearly and over time. While time-varying and non-linear effects have been used in relational event modeling, we take this further by modeling joint time-varying and non-linear effects using tensor product smooths.
We validate our methodology on both synthetic and empirical data. In particular, we use RHEMs to study how patterns of scientific collaboration and impact evolve over time. Our approach provides deeper insights into the dynamic factors driving relational hyper-events, allowing us to evaluate potential non-monotonic patterns that cannot be identified using linear models.
|
[
{
"version": "v1",
"created": "Fri, 05 Sep 2025 17:55:29 GMT"
},
{
"version": "v2",
"created": "Mon, 08 Sep 2025 06:53:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Boschi",
"Martina",
""
],
[
"Lerner",
"Jürgen",
""
],
[
"Wit",
"Ernst C.",
""
]
] |
2509.05298
|
Xianghan Wang
|
Rui Xi, Xianghan Wang
|
Livia: An Emotion-Aware AR Companion Powered by Modular AI Agents and Progressive Memory Compression
|
Accepted to the Proceedings of the 2025 International Conference on Artificial Intelligence and Virtual Reality (AIVR 2025). \c{opyright} 2025 Springer. This is the author-accepted manuscript. Rui Xi and Xianghan Wang contributed equally to this work. The final version will be available via SpringerLink
| null | null | null |
cs.HC cs.AI cs.MM
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Loneliness and social isolation pose significant emotional and health challenges, prompting the development of technology-based solutions for companionship and emotional support. This paper introduces Livia, an emotion-aware augmented reality (AR) companion app designed to provide personalized emotional support by combining modular artificial intelligence (AI) agents, multimodal affective computing, progressive memory compression, and AR driven embodied interaction. Livia employs a modular AI architecture with specialized agents responsible for emotion analysis, dialogue generation, memory management, and behavioral orchestration, ensuring robust and adaptive interactions. Two novel algorithms-Temporal Binary Compression (TBC) and Dynamic Importance Memory Filter (DIMF)-effectively manage and prioritize long-term memory, significantly reducing storage requirements while retaining critical context. Our multimodal emotion detection approach achieves high accuracy, enhancing proactive and empathetic engagement. User evaluations demonstrated increased emotional bonds, improved satisfaction, and statistically significant reductions in loneliness. Users particularly valued Livia's adaptive personality evolution and realistic AR embodiment. Future research directions include expanding gesture and tactile interactions, supporting multi-user experiences, and exploring customized hardware implementations.
|
[
{
"version": "v1",
"created": "Tue, 12 Aug 2025 14:07:22 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Xi",
"Rui",
""
],
[
"Wang",
"Xianghan",
""
]
] |
2509.05299
|
Paul Poncet
|
Paul Poncet
|
Order-generation in posets and convolution of closure operators
|
50 pages
| null | null | null |
math.CO math.FA math.GN
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Motivated by the Hofmann--Lawson theorem, which states that every continuous lattice is inf-generated by its irreducible elements, we explore how to represent posets by extreme points with respect to a closure operator. For this purpose, we introduce the convolution product of closure operators, and prove that the Krein--Milman property can be transferred from one collection of subsets to another by convolution. This result underpins two new representation theorems of topological flavor, which generalize existing ones, even in posets lacking lattice or semilattice structures. We also prove a third representation theorem: given a poset equipped with a closure operator $\mathfrak{c}$ with adequate properties, we show that the set of kit points, defined as an extension of compact points, has the Krein--Milman property with respect to the convolution product of $\mathfrak{c}$ with the dual Alexandrov operator $\uparrow\!\! \cdot$; moreover, every kit point is sup-generated by a unique antichain of compact points, finite if $\mathfrak{c}$ is finitary.
|
[
{
"version": "v1",
"created": "Sat, 16 Aug 2025 15:08:39 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Poncet",
"Paul",
""
]
] |
2509.05301
|
Ariel Pedrano
|
Ariel C. Pedrano and Rolando N. Paluga
|
On 2-Movable Domination in the Join and Corona of Graphs
|
This article is already published
|
Volume 42 Number 2 (2025)
|
10.17654/0974165825007
| null |
math.GM
|
http://creativecommons.org/licenses/by/4.0/
|
Let $G$ be a connected graph. A non-empty $S\subseteq V(G)$ is a $2$-movable dominating set of $G$ if $S$ is a dominating set and for every pair $x,y \in S$, $S\backslash \{x, y\}$ is a dominating set in $G$, or there exist $u, v \in V(G) \backslash S$ such that $u$ and $v$ are adjacent to $x$ and $y$, respectively, and $(S \backslash \{x,y\}) \cup \{u,v\}$ is a dominating set in $G$. The $2$-movable domination number of $G$, denoted by $\gamma_{m}^{2}(G)$, is the minimum cardinality of a 2-movable dominating set of $G$. A 2-movable dominating set with cardinality equal to $\gamma_{m}^{2}(G)$ is called $\gamma_{m}^{2}$-set of $G$.
This paper present the 2-movable domination number in the corona and join of graphs.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 04:40:16 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Pedrano",
"Ariel C.",
""
],
[
"Paluga",
"Rolando N.",
""
]
] |
2509.05302
|
Alexis Molina
|
Ra\'ul Mi\~n\'an, Carles Perez-Lopez, Javier Iglesias, \'Alvaro Ciudad, Alexis Molina
|
Sesame: Opening the door to protein pockets
|
Published at the Proceedings of the 2nd Workshop on Generative and Experimental Perspectives for Biomolecular Design. ICLR 2025
| null | null | null |
q-bio.BM cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Molecular docking is a cornerstone of drug discovery, relying on high-resolution ligand-bound structures to achieve accurate predictions. However, obtaining these structures is often costly and time-intensive, limiting their availability. In contrast, ligand-free structures are more accessible but suffer from reduced docking performance due to pocket geometries being less suited for ligand accommodation in apo structures. Traditional methods for artificially inducing these conformations, such as molecular dynamics simulations, are computationally expensive. In this work, we introduce Sesame, a generative model designed to predict this conformational change efficiently. By generating geometries better suited for ligand accommodation at a fraction of the computational cost, Sesame aims to provide a scalable solution for improving virtual screening workflows.
|
[
{
"version": "v1",
"created": "Thu, 21 Aug 2025 12:22:56 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Miñán",
"Raúl",
""
],
[
"Perez-Lopez",
"Carles",
""
],
[
"Iglesias",
"Javier",
""
],
[
"Ciudad",
"Álvaro",
""
],
[
"Molina",
"Alexis",
""
]
] |
2509.05303
|
Sam Davidson
|
Sam Davidson, Li Sun, Bhavana Bhasker, Laurent Callot, Anoop Deoras
|
Multi-IaC-Eval: Benchmarking Cloud Infrastructure as Code Across Multiple Formats
| null | null | null | null |
cs.DC cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Infrastructure as Code (IaC) is fundamental to modern cloud computing, enabling teams to define and manage infrastructure through machine-readable configuration files. However, different cloud service providers utilize diverse IaC formats. The lack of a standardized format requires cloud architects to be proficient in multiple IaC languages, adding complexity to cloud deployment. While Large Language Models (LLMs) show promise in automating IaC creation and maintenance, progress has been limited by the lack of comprehensive benchmarks across multiple IaC formats. We present Multi-IaC-Bench, a novel benchmark dataset for evaluating LLM-based IaC generation and mutation across AWS CloudFormation, Terraform, and Cloud Development Kit (CDK) formats. The dataset consists of triplets containing initial IaC templates, natural language modification requests, and corresponding updated templates, created through a synthetic data generation pipeline with rigorous validation. We evaluate several state-of-the-art LLMs on Multi-IaC-Bench, demonstrating that while modern LLMs can achieve high success rates (>95%) in generating syntactically valid IaC across formats, significant challenges remain in semantic alignment and handling complex infrastructure patterns. Our ablation studies highlight the importance of prompt engineering and retry mechanisms in successful IaC generation. We release Multi-IaC-Bench to facilitate further research in AI-assisted infrastructure management and establish standardized evaluation metrics for this crucial domain.
|
[
{
"version": "v1",
"created": "Thu, 21 Aug 2025 22:37:18 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Davidson",
"Sam",
""
],
[
"Sun",
"Li",
""
],
[
"Bhasker",
"Bhavana",
""
],
[
"Callot",
"Laurent",
""
],
[
"Deoras",
"Anoop",
""
]
] |
2509.05304
|
Itai Zilberstein
|
Steve Chien, Itai Zilberstein, Alberto Candela, David Rijlaarsdam, Tom Hendrix, Aubrey Dunne, Aragon Oriol, Miquel Juan Puig
|
Flight of Dynamic Targeting on the CogniSAT-6 Spacecraft
| null |
International Symposium on Artificial Intelligence, Robotics, and Automation for Space, November 2024
| null | null |
eess.SY astro-ph.IM cs.SY
|
http://creativecommons.org/licenses/by/4.0/
|
Dynamic targeting (DT) is a spacecraft autonomy concept in which sensor data is acquired and rapidly analyzed and used to drive subsequent observation. We describe the low Earth orbit application of this approach in which lookahead imagery is analyzed to detect clouds, thermal anomalies, or land use cases to drive higher quality near nadir imaging. Use cases for such a capability include: cloud avoidance, storm hunting, search for planetary boundary layer events, plume study, and beyond. The DT concept requires a lookahead sensor or agility to use a primary sensor in such a mode, edge computing to analyze images rapidly onboard, and a primary followup sensor. Additionally, an inter-satellite or low latency communications link can be leveraged for cross platform tasking. We describe implementation in progress to fly DT in early 2025 on the CogniSAT-6 (Ubotica/Open Cosmos) spacecraft that launched in March 2024 on the SpaceX Transporter-10 launch.
|
[
{
"version": "v1",
"created": "Fri, 22 Aug 2025 02:32:42 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chien",
"Steve",
""
],
[
"Zilberstein",
"Itai",
""
],
[
"Candela",
"Alberto",
""
],
[
"Rijlaarsdam",
"David",
""
],
[
"Hendrix",
"Tom",
""
],
[
"Dunne",
"Aubrey",
""
],
[
"Oriol",
"Aragon",
""
],
[
"Puig",
"Miquel Juan",
""
]
] |
2509.05305
|
Xu Zhiping
|
Yingjie Zhao and Yicheng Song and Fan Xu and Zhiping Xu
|
Predicting Brain Morphogenesis via Physics-Transfer Learning
| null | null | null | null |
q-bio.NC cs.LG nlin.PS
|
http://creativecommons.org/licenses/by/4.0/
|
Brain morphology is shaped by genetic and mechanical factors and is linked to biological development and diseases. Its fractal-like features, regional anisotropy, and complex curvature distributions hinder quantitative insights in medical inspections. Recognizing that the underlying elastic instability and bifurcation share the same physics as simple geometries such as spheres and ellipses, we developed a physics-transfer learning framework to address the geometrical complexity. To overcome the challenge of data scarcity, we constructed a digital library of high-fidelity continuum mechanics modeling that both describes and predicts the developmental processes of brain growth and disease. The physics of nonlinear elasticity from simple geometries is embedded into a neural network and applied to brain models. This physics-transfer approach demonstrates remarkable performance in feature characterization and morphogenesis prediction, highlighting the pivotal role of localized deformation in dominating over the background geometry. The data-driven framework also provides a library of reduced-dimensional evolutionary representations that capture the essential physics of the highly folded cerebral cortex. Validation through medical images and domain expertise underscores the deployment of digital-twin technology in comprehending the morphological complexity of the brain.
|
[
{
"version": "v1",
"created": "Fri, 22 Aug 2025 09:45:13 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Zhao",
"Yingjie",
""
],
[
"Song",
"Yicheng",
""
],
[
"Xu",
"Fan",
""
],
[
"Xu",
"Zhiping",
""
]
] |
2509.05306
|
Enis Karaarslan Dr.
|
Enis Karaarslan, Esin G\"uler, Efe Emir Y\"uce, Cagatay Coban
|
Towards Log Analysis with AI Agents: Cowrie Case Study
| null | null | null | null |
cs.CR cs.AI cs.MA
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
The scarcity of real-world attack data significantly hinders progress in cybersecurity research and education. Although honeypots like Cowrie effectively collect live threat intelligence, they generate overwhelming volumes of unstructured and heterogeneous logs, rendering manual analysis impractical. As a first step in our project on secure and efficient AI automation, this study explores the use of AI agents for automated log analysis. We present a lightweight and automated approach to process Cowrie honeypot logs. Our approach leverages AI agents to intelligently parse, summarize, and extract insights from raw data, while also considering the security implications of deploying such an autonomous system. Preliminary results demonstrate the pipeline's effectiveness in reducing manual effort and identifying attack patterns, paving the way for more advanced autonomous cybersecurity analysis in future work.
|
[
{
"version": "v1",
"created": "Fri, 22 Aug 2025 16:50:59 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Karaarslan",
"Enis",
""
],
[
"Güler",
"Esin",
""
],
[
"Yüce",
"Efe Emir",
""
],
[
"Coban",
"Cagatay",
""
]
] |
2509.05307
|
Sachin Chhabra
|
Sachin Chhabra, Hemanth Venkateswara and Baoxin Li
|
Label Smoothing++: Enhanced Label Regularization for Training Neural Networks
|
Published in British Machine Vision Conference (BMVC), 2024
|
Proc. British Machine Vision Conference (BMVC), 2024
| null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Training neural networks with one-hot target labels often results in overconfidence and overfitting. Label smoothing addresses this issue by perturbing the one-hot target labels by adding a uniform probability vector to create a regularized label. Although label smoothing improves the network's generalization ability, it assigns equal importance to all the non-target classes, which destroys the inter-class relationships. In this paper, we propose a novel label regularization training strategy called Label Smoothing++, which assigns non-zero probabilities to non-target classes and accounts for their inter-class relationships. Our approach uses a fixed label for the target class while enabling the network to learn the labels associated with non-target classes. Through extensive experiments on multiple datasets, we demonstrate how Label Smoothing++ mitigates overconfident predictions while promoting inter-class relationships and generalization capabilities.
|
[
{
"version": "v1",
"created": "Fri, 22 Aug 2025 23:11:43 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chhabra",
"Sachin",
""
],
[
"Venkateswara",
"Hemanth",
""
],
[
"Li",
"Baoxin",
""
]
] |
2509.05308
|
Francesco Mainardi
|
Luiz Roberto Evangelista and Francesco Mainardi
|
The Tautochrone of Huygens and Abel: From Constructive Geometry to Fractional Calculus
|
24 pages, 3 figures
| null | null | null |
math.HO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this paper, we explore the connections between Christiaan Huygens and Niels Henrik Abel through the tautochrone problem. The problem -- determining the curve along which a particle descends under gravity in the same time, regardless of its starting point -- has been a central topic at the intersection of physics, geometry, and analysis. Though these two major figures are separated by nearly two centuries, they approached the problem in radically different ways. While Huygens proposed a physical solution based on geometric construction, Abel approached the problem within the analytic framework of integral equations, employing a procedure that can be seen as anticipating and paving the way for the development of differential calculus of arbitrary order. This contrast highlights a broader historical narrative: the transformation of mathematical thinking from constructive geometry to abstract analysis.
|
[
{
"version": "v1",
"created": "Mon, 25 Aug 2025 08:32:30 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Evangelista",
"Luiz Roberto",
""
],
[
"Mainardi",
"Francesco",
""
]
] |
2509.05309
|
Wei Hu
|
Xiangyu Liu and Haodi Lei and Yi Liu and Yang Liu and Wei Hu
|
ProtSAE: Disentangling and Interpreting Protein Language Models via Semantically-Guided Sparse Autoencoders
| null | null | null | null |
q-bio.QM cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Sparse Autoencoder (SAE) has emerged as a powerful tool for mechanistic interpretability of large language models. Recent works apply SAE to protein language models (PLMs), aiming to extract and analyze biologically meaningful features from their latent spaces. However, SAE suffers from semantic entanglement, where individual neurons often mix multiple nonlinear concepts, making it difficult to reliably interpret or manipulate model behaviors. In this paper, we propose a semantically-guided SAE, called ProtSAE. Unlike existing SAE which requires annotation datasets to filter and interpret activations, we guide semantic disentanglement during training using both annotation datasets and domain knowledge to mitigate the effects of entangled attributes. We design interpretability experiments showing that ProtSAE learns more biologically relevant and interpretable hidden features compared to previous methods. Performance analyses further demonstrate that ProtSAE maintains high reconstruction fidelity while achieving better results in interpretable probing. We also show the potential of ProtSAE in steering PLMs for downstream generation tasks.
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 11:20:31 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Liu",
"Xiangyu",
""
],
[
"Lei",
"Haodi",
""
],
[
"Liu",
"Yi",
""
],
[
"Liu",
"Yang",
""
],
[
"Hu",
"Wei",
""
]
] |
2509.05310
|
Chang Ge
|
Sabina Tomkins, Chang Ge, David Rothschild
|
How candidates evoke identity and issues on TikTok
| null | null | null | null |
cs.CY cs.SI
|
http://creativecommons.org/licenses/by/4.0/
|
Social media platforms are increasingly central to campaign communication, with both paid (advertising) and earned (organic) posts used for fundraising, mobilization, and persuasion. TikTok, and other short-form video platforms, with its short-video format and content-driven algorithms, demand unique content. We examine the final six months before the 2024 US Presidential Election to understand how major campaigns used TikTok. We frame our analysis around two political science theories. The first is the expressive (identity) model, where voters are motivated by their group memberships and candidates appeal to those identities. Alternatively, the instrumental (issues) model argues voters align with politicians advocating their key issues. We also examine how often candidates attacked opponents, reflecting literature showing attacks are common in politics. We combine two datasets: posts from the Harris and Trump campaigns on TikTok (July-November 2024) and a two-wave 2022 survey of around 1,000 respondents. Results show Trump more often disparaged Harris and emphasized identities and issues distinguishing Republicans, while Harris more often highlighted Democratic identities and valued issues. Although issues predict party ID, both candidates referenced identities more (34 percent of posts) than issues (25 percent), with most posts mentioning neither (55 percent).
|
[
{
"version": "v1",
"created": "Tue, 26 Aug 2025 13:27:42 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Tomkins",
"Sabina",
""
],
[
"Ge",
"Chang",
""
],
[
"Rothschild",
"David",
""
]
] |
2509.05311
|
Konur Tholl
|
Konur Tholl, Fran\c{c}ois Rivest, Mariam El Mezouar and Ranwa Al Mallah
|
Large Language Model Integration with Reinforcement Learning to Augment Decision-Making in Autonomous Cyber Operations
| null | null | null | null |
cs.CR cs.AI cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Reinforcement Learning (RL) has shown great potential for autonomous decision-making in the cybersecurity domain, enabling agents to learn through direct environment interaction. However, RL agents in Autonomous Cyber Operations (ACO) typically learn from scratch, requiring them to execute undesirable actions to learn their consequences. In this study, we integrate external knowledge in the form of a Large Language Model (LLM) pretrained on cybersecurity data that our RL agent can directly leverage to make informed decisions. By guiding initial training with an LLM, we improve baseline performance and reduce the need for exploratory actions with obviously negative outcomes. We evaluate our LLM-integrated approach in a simulated cybersecurity environment, and demonstrate that our guided agent achieves over 2x higher rewards during early training and converges to a favorable policy approximately 4,500 episodes faster than the baseline.
|
[
{
"version": "v1",
"created": "Thu, 28 Aug 2025 23:52:03 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Tholl",
"Konur",
""
],
[
"Rivest",
"François",
""
],
[
"Mezouar",
"Mariam El",
""
],
[
"Mallah",
"Ranwa Al",
""
]
] |
2509.05312
|
Xinghjua Cui
|
Cui Xinghua, Wang Haoyang, and Peng Zhifeng
|
The trace formula of GL(3)
| null | null | null | null |
math.RT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Trace formula is an important method to study the Langlands program. In this paper, we will give the explicit coarse trace formula of $\GL(3)$. In general case, Arthur applies the truncation operator on the two sides of trace formula such that they are convergent. In our case, we will prove that the divergent terms of the two sides of the trace formula of $\GL(3)$ are equal. We also obtain the explicit formula for ramified orbits of the geometric side of $\GL(3)$, and show it can be expressed as a limit of distributions of the unramified orbits.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 05:26:37 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Xinghua",
"Cui",
""
],
[
"Haoyang",
"Wang",
""
],
[
"Zhifeng",
"Peng",
""
]
] |
2509.05313
|
Haoyang Wang
|
Haoyang Wang, Xinghua Cui, Zhifeng Peng
|
The coarse trace formula of GL(4)
|
69 pages
| null | null | null |
math.RT math.NT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Trace formula is an important method to study the Langlands program. Arthur obtains the existence of stable trace formula for connected reductive group. In this paper, we will give the explicit coarse trace formula of GL(4). In general case, Arthur applies the truncation operator on the two sides of trace formula, which is convergent. In our case, we will prove that the divergent terms of the two sides of the trace formula of GL(4)$ are equal. We also obtain the explicit formula for ramified orbits of the geometric side of trace formula of GL(4).
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 08:29:15 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Wang",
"Haoyang",
""
],
[
"Cui",
"Xinghua",
""
],
[
"Peng",
"Zhifeng",
""
]
] |
2509.05314
|
Ying Li
|
Ying Li, Xiaobao Wei, Xiaowei Chi, Yuming Li, Zhongyu Zhao, Hao Wang, Ningning Ma, Ming Lu, Shanghang Zhang
|
ManipDreamer3D : Synthesizing Plausible Robotic Manipulation Video with Occupancy-aware 3D Trajectory
|
8pages; 7figures; 4 tables
| null | null | null |
cs.RO cs.AI cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Data scarcity continues to be a major challenge in the field of robotic manipulation. Although diffusion models provide a promising solution for generating robotic manipulation videos, existing methods largely depend on 2D trajectories, which inherently face issues with 3D spatial ambiguity. In this work, we present a novel framework named ManipDreamer3D for generating plausible 3D-aware robotic manipulation videos from the input image and the text instruction. Our method combines 3D trajectory planning with a reconstructed 3D occupancy map created from a third-person perspective, along with a novel trajectory-to-video diffusion model. Specifically, ManipDreamer3D first reconstructs the 3D occupancy representation from the input image and then computes an optimized 3D end-effector trajectory, minimizing path length while avoiding collisions. Next, we employ a latent editing technique to create video sequences from the initial image latent and the optimized 3D trajectory. This process conditions our specially trained trajectory-to-video diffusion model to produce robotic pick-and-place videos. Our method generates robotic videos with autonomously planned plausible 3D trajectories, significantly reducing human intervention requirements. Experimental results demonstrate superior visual quality compared to existing methods.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 10:39:06 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Li",
"Ying",
""
],
[
"Wei",
"Xiaobao",
""
],
[
"Chi",
"Xiaowei",
""
],
[
"Li",
"Yuming",
""
],
[
"Zhao",
"Zhongyu",
""
],
[
"Wang",
"Hao",
""
],
[
"Ma",
"Ningning",
""
],
[
"Lu",
"Ming",
""
],
[
"Zhang",
"Shanghang",
""
]
] |
2509.05315
|
Angelos Amanatiadis
|
Petros Loukas, David Bassir, Savvas Chatzichristofis, Angelos Amanatiadis
|
Evaluation of Large Language Models for Anomaly Detection in Autonomous Vehicles
| null | null | null | null |
cs.RO cs.AI cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
The rapid evolution of large language models (LLMs) has pushed their boundaries to many applications in various domains. Recently, the research community has started to evaluate their potential adoption in autonomous vehicles and especially as complementary modules in the perception and planning software stacks. However, their evaluation is limited in synthetic datasets or manually driving datasets without the ground truth knowledge and more precisely, how the current perception and planning algorithms would perform in the cases under evaluation. For this reason, this work evaluates LLMs on real-world edge cases where current autonomous vehicles have been proven to fail. The proposed architecture consists of an open vocabulary object detector coupled with prompt engineering and large language model contextual reasoning. We evaluate several state-of-the-art models against real edge cases and provide qualitative comparison results along with a discussion on the findings for the potential application of LLMs as anomaly detectors in autonomous vehicles.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 13:05:13 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Loukas",
"Petros",
""
],
[
"Bassir",
"David",
""
],
[
"Chatzichristofis",
"Savvas",
""
],
[
"Amanatiadis",
"Angelos",
""
]
] |
2509.05316
|
Praveen Bushipaka
|
Praveen Bushipaka, Lucia Passaro and Tommaso Cucinotta
|
Standard vs. Modular Sampling: Best Practices for Reliable LLM Unlearning
| null | null | null | null |
cs.LG cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
A conventional LLM Unlearning setting consists of two subsets -"forget" and "retain", with the objectives of removing the undesired knowledge from the forget set while preserving the remaining knowledge from the retain. In privacy-focused unlearning research, a retain set is often further divided into neighbor sets, containing either directly or indirectly connected to the forget targets; and augmented by a general-knowledge set. A common practice in existing benchmarks is to employ only a single neighbor set, with general knowledge which fails to reflect the real-world data complexities and relationships. LLM Unlearning typically involves 1:1 sampling or cyclic iteration sampling. However, the efficacy and stability of these de facto standards have not been critically examined. In this study, we systematically evaluate these common practices. Our findings reveal that relying on a single neighbor set is suboptimal and that a standard sampling approach can obscure performance trade-offs. Based on this analysis, we propose and validate an initial set of best practices: (1) Incorporation of diverse neighbor sets to balance forget efficacy and model utility, (2) Standard 1:1 sampling methods are inefficient and yield poor results, (3) Our proposed Modular Entity-Level Unlearning (MELU) strategy as an alternative to cyclic sampling. We demonstrate that this modular approach, combined with robust algorithms, provides a clear and stable path towards effective unlearning.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 19:25:52 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Bushipaka",
"Praveen",
""
],
[
"Passaro",
"Lucia",
""
],
[
"Cucinotta",
"Tommaso",
""
]
] |
2509.05317
|
Isac Holm
|
Isac Holm
|
VILOD: A Visual Interactive Labeling Tool for Object Detection
|
Master's project
| null | null | null |
cs.CV cs.AI cs.HC cs.LG
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The advancement of Object Detection (OD) using Deep Learning (DL) is often hindered by the significant challenge of acquiring large, accurately labeled datasets, a process that is time-consuming and expensive. While techniques like Active Learning (AL) can reduce annotation effort by intelligently querying informative samples, they often lack transparency, limit the strategic insight of human experts, and may overlook informative samples not aligned with an employed query strategy. To mitigate these issues, Human-in-the-Loop (HITL) approaches integrating human intelligence and intuition throughout the machine learning life-cycle have gained traction. Leveraging Visual Analytics (VA), effective interfaces can be created to facilitate this human-AI collaboration. This thesis explores the intersection of these fields by developing and investigating "VILOD: A Visual Interactive Labeling tool for Object Detection". VILOD utilizes components such as a t-SNE projection of image features, together with uncertainty heatmaps and model state views. Enabling users to explore data, interpret model states, AL suggestions, and implement diverse sample selection strategies within an iterative HITL workflow for OD. An empirical investigation using comparative use cases demonstrated how VILOD, through its interactive visualizations, facilitates the implementation of distinct labeling strategies by making the model's state and dataset characteristics more interpretable (RQ1). The study showed that different visually-guided labeling strategies employed within VILOD result in competitive OD performance trajectories compared to an automated uncertainty sampling AL baseline (RQ2). This work contributes a novel tool and empirical insight into making the HITL-AL workflow for OD annotation more transparent, manageable, and potentially more effective.
|
[
{
"version": "v1",
"created": "Fri, 29 Aug 2025 19:27:10 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Holm",
"Isac",
""
]
] |
2509.05318
|
Zuquan Peng
|
Zuquan Peng and Jianming Fu and Lixin Zou and Li Zheng and Yanzhen Ren and Guojun Peng
|
Backdoor Samples Detection Based on Perturbation Discrepancy Consistency in Pre-trained Language Models
|
13 pages, 9 figures, 8 tables, journal
|
Neural Networks 193(2026) 108025
|
10.1016/j.neunet.2025.108025
| null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
The use of unvetted third-party and internet data renders pre-trained models susceptible to backdoor attacks. Detecting backdoor samples is critical to prevent backdoor activation during inference or injection during training. However, existing detection methods often require the defender to have access to the poisoned models, extra clean samples, or significant computational resources to detect backdoor samples, limiting their practicality. To address this limitation, we propose a backdoor sample detection method based on perturbatio\textbf{N} discr\textbf{E}pancy consis\textbf{T}ency \textbf{E}valuation (\NETE). This is a novel detection method that can be used both pre-training and post-training phases. In the detection process, it only requires an off-the-shelf pre-trained model to compute the log probability of samples and an automated function based on a mask-filling strategy to generate perturbations. Our method is based on the interesting phenomenon that the change in perturbation discrepancy for backdoor samples is smaller than that for clean samples. Based on this phenomenon, we use curvature to measure the discrepancy in log probabilities between different perturbed samples and input samples, thereby evaluating the consistency of the perturbation discrepancy to determine whether the input sample is a backdoor sample. Experiments conducted on four typical backdoor attacks and five types of large language model backdoor attacks demonstrate that our detection strategy outperforms existing zero-shot black-box detection methods.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 06:35:32 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Peng",
"Zuquan",
""
],
[
"Fu",
"Jianming",
""
],
[
"Zou",
"Lixin",
""
],
[
"Zheng",
"Li",
""
],
[
"Ren",
"Yanzhen",
""
],
[
"Peng",
"Guojun",
""
]
] |
2509.05319
|
ZhengDa Li
|
Zhengda Li
|
Context-Aware Knowledge Distillation with Adaptive Weighting for Image Classification
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Knowledge distillation (KD) is a widely used technique to transfer knowledge from a large teacher network to a smaller student model. Traditional KD uses a fixed balancing factor alpha as a hyperparameter to combine the hard-label cross-entropy loss with the soft-label distillation loss. However, a static alpha is suboptimal because the optimal trade-off between hard and soft supervision can vary during training.
In this work, we propose an Adaptive Knowledge Distillation (AKD) framework. First we try to make alpha as learnable parameter that can be automatically learned and optimized during training. Then we introduce a formula to reflect the gap between the student and the teacher to compute alpha dynamically, guided by student-teacher discrepancies, and further introduce a Context-Aware Module (CAM) using MLP + Attention to adaptively reweight class-wise teacher outputs. Experiments on CIFAR-10 with ResNet-50 as teacher and ResNet-18 as student demonstrate that our approach achieves superior accuracy compared to fixed-weight KD baselines, and yields more stable convergence.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 09:18:46 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Li",
"Zhengda",
""
]
] |
2509.05320
|
Amine Abouaomar
|
Ikhlasse Badidi, Nouhaila El Khiyaoui, Aya Riany, Badr Ben Elallid, Amine Abouaomar
|
Privacy-Preserving Offloading for Large Language Models in 6G Vehicular Networks
|
7 pages, 6 figures, 1 algorithm, 5 equations
| null | null | null |
cs.CR cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
The integration of Large Language Models (LLMs) in 6G vehicular networks promises unprecedented advancements in intelligent transportation systems. However, offloading LLM computations from vehicles to edge infrastructure poses significant privacy risks, potentially exposing sensitive user data. This paper presents a novel privacy-preserving offloading framework for LLM-integrated vehicular networks. We introduce a hybrid approach combining federated learning (FL) and differential privacy (DP) techniques to protect user data while maintaining LLM performance. Our framework includes a privacy-aware task partitioning algorithm that optimizes the trade-off between local and edge computation, considering both privacy constraints and system efficiency. We also propose a secure communication protocol for transmitting model updates and aggregating results across the network. Experimental results demonstrate that our approach achieves 75\% global accuracy with only a 2-3\% reduction compared to non-privacy-preserving methods, while maintaining DP guarantees with an optimal privacy budget of $\varepsilon = 0.8$. The framework shows stable communication overhead of approximately 2.1MB per round with computation comprising over 90\% of total processing time, validating its efficiency for resource-constrained vehicular environments.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 10:08:28 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Badidi",
"Ikhlasse",
""
],
[
"Khiyaoui",
"Nouhaila El",
""
],
[
"Riany",
"Aya",
""
],
[
"Elallid",
"Badr Ben",
""
],
[
"Abouaomar",
"Amine",
""
]
] |
2509.05321
|
Yunfei Guo
|
Yunfei Guo, Tao Zhang, Wu Huang, Yao Song
|
A Dataset Generation Scheme Based on Video2EEG-SPGN-Diffusion for SEED-VD
| null | null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
This paper introduces an open-source framework, Video2EEG-SPGN-Diffusion, that leverages the SEED-VD dataset to generate a multimodal dataset of EEG signals conditioned on video stimuli. Additionally, we disclose an engineering pipeline for aligning video and EEG data pairs, facilitating the training of multimodal large models with EEG alignment capabilities. Personalized EEG signals are generated using a self-play graph network (SPGN) integrated with a diffusion model. As a major contribution, we release a new dataset comprising over 1000 samples of SEED-VD video stimuli paired with generated 62-channel EEG signals at 200 Hz and emotion labels, enabling video-EEG alignment and advancing multimodal research. This framework offers novel tools for emotion analysis, data augmentation, and brain-computer interface applications, with substantial research and engineering significance.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 10:52:17 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Guo",
"Yunfei",
""
],
[
"Zhang",
"Tao",
""
],
[
"Huang",
"Wu",
""
],
[
"Song",
"Yao",
""
]
] |
2509.05322
|
Areejit Samal
|
Pavithra Elumalai, Sudharsan Vijayaraghavan, Madhumita Mondal, Areejit Samal
|
Application of discrete Ricci curvature in pruning randomly wired neural networks: A case study with chest x-ray classification of COVID-19
|
21 pages, 4 figures, 9 tables
| null | null | null |
cs.CV cs.LG cs.SI physics.comp-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Randomly Wired Neural Networks (RWNNs) serve as a valuable testbed for investigating the impact of network topology in deep learning by capturing how different connectivity patterns impact both learning efficiency and model performance. At the same time, they provide a natural framework for exploring edge-centric network measures as tools for pruning and optimization. In this study, we investigate three edge-centric network measures: Forman-Ricci curvature (FRC), Ollivier-Ricci curvature (ORC), and edge betweenness centrality (EBC), to compress RWNNs by selectively retaining important synapses (or edges) while pruning the rest. As a baseline, RWNNs are trained for COVID-19 chest x-ray image classification, aiming to reduce network complexity while preserving performance in terms of accuracy, specificity, and sensitivity. We extend prior work on pruning RWNN using ORC by incorporating two additional edge-centric measures, FRC and EBC, across three network generators: Erd\"{o}s-R\'{e}nyi (ER) model, Watts-Strogatz (WS) model, and Barab\'{a}si-Albert (BA) model. We provide a comparative analysis of the pruning performance of the three measures in terms of compression ratio and theoretical speedup. A central focus of our study is to evaluate whether FRC, which is computationally more efficient than ORC, can achieve comparable pruning effectiveness. Along with performance evaluation, we further investigate the structural properties of the pruned networks through modularity and global efficiency, offering insights into the trade-off between modular segregation and network efficiency in compressed RWNNs. Our results provide initial evidence that FRC-based pruning can effectively simplify RWNNs, offering significant computational advantages while maintaining performance comparable to ORC.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 18:50:22 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Elumalai",
"Pavithra",
""
],
[
"Vijayaraghavan",
"Sudharsan",
""
],
[
"Mondal",
"Madhumita",
""
],
[
"Samal",
"Areejit",
""
]
] |
2509.05324
|
Rongqian Chen
|
Rongqian Chen, Shu Hong, Rifatul Islam, Mahdi Imani, G. Gary Tan, Tian Lan
|
Perception Graph for Cognitive Attack Reasoning in Augmented Reality
|
Accepted by ACM MobiHoc XR Security workshop 2025
| null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Augmented reality (AR) systems are increasingly deployed in tactical environments, but their reliance on seamless human-computer interaction makes them vulnerable to cognitive attacks that manipulate a user's perception and severely compromise user decision-making. To address this challenge, we introduce the Perception Graph, a novel model designed to reason about human perception within these systems. Our model operates by first mimicking the human process of interpreting key information from an MR environment and then representing the outcomes using a semantically meaningful structure. We demonstrate how the model can compute a quantitative score that reflects the level of perception distortion, providing a robust and measurable method for detecting and analyzing the effects of such cognitive attacks.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 20:48:32 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chen",
"Rongqian",
""
],
[
"Hong",
"Shu",
""
],
[
"Islam",
"Rifatul",
""
],
[
"Imani",
"Mahdi",
""
],
[
"Tan",
"G. Gary",
""
],
[
"Lan",
"Tian",
""
]
] |
2509.05325
|
Liming Xu Dr
|
Liming Xu and Yunbo Long and Alexandra Brintrup
|
SynDelay: A Synthetic Dataset for Delivery Delay Prediction
|
This paper incldues 1 figure and 2 tables
| null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Artificial intelligence (AI) is transforming supply chain management, yet progress in predictive tasks -- such as delivery delay prediction -- remains constrained by the scarcity of high-quality, openly available datasets. Existing datasets are often proprietary, small, or inconsistently maintained, hindering reproducibility and benchmarking. We present SynDelay, a synthetic dataset designed for delivery delay prediction. Generated using an advanced generative model trained on real-world data, SynDelay preserves realistic delivery patterns while ensuring privacy. Although not entirely free of noise or inconsistencies, it provides a challenging and practical testbed for advancing predictive modelling. To support adoption, we provide baseline results and evaluation metrics as initial benchmarks, serving as reference points rather than state-of-the-art claims. SynDelay is publicly available through the Supply Chain Data Hub, an open initiative promoting dataset sharing and benchmarking in supply chain AI. We encourage the community to contribute datasets, models, and evaluation practices to advance research in this area. All code is openly accessible at https://supplychaindatahub.org.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 21:54:37 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Xu",
"Liming",
""
],
[
"Long",
"Yunbo",
""
],
[
"Brintrup",
"Alexandra",
""
]
] |
2509.05326
|
Logan Nye
|
Logan Nye
|
Zero-Knowledge Proofs in Sublinear Space
|
21 pages
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Modern zero-knowledge proof (ZKP) systems, essential for privacy and verifiable computation, suffer from a fundamental limitation: the prover typically uses memory that scales linearly with the computation's trace length T, making them impractical for resource-constrained devices and prohibitively expensive for large-scale tasks. This paper overcomes this barrier by constructing, to our knowledge, the first sublinear-space ZKP prover. Our core contribution is an equivalence that reframes proof generation as an instance of the classic Tree Evaluation problem. Leveraging a recent space-efficient tree-evaluation algorithm, we design a streaming prover that assembles the proof without ever materializing the full execution trace. The approach reduces prover memory from linear in T to O(sqrt(T)) (up to O(log T) lower-order terms) while preserving proof size, verifier time, and the transcript/security guarantees of the underlying system. This enables a shift from specialized, server-bound proving to on-device proving, opening applications in decentralized systems, on-device machine learning, and privacy-preserving technologies.
|
[
{
"version": "v1",
"created": "Sat, 30 Aug 2025 23:22:54 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Nye",
"Logan",
""
]
] |
2509.05327
|
Ping Guo
|
Guanzhong Hu, Wenpan Li, Rujing Zha, Ping Guo
|
Layer-Wise Anomaly Detection in Directed Energy Deposition using High-Fidelity Fringe Projection Profilometry
|
26 pages, 15 figures
| null | null | null |
physics.optics cs.CV
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Directed energy deposition (DED), a metal additive manufacturing process, is highly susceptible to process-induced defects such as geometric deviations, lack of fusion, and poor surface finish. This work presents a build-height-synchronized fringe projection system for in-situ, layer-wise surface reconstruction of laser-DED components, achieving a reconstruction accuracy of ${\pm}$46 ${\mu}$m. From the reconstructed 3D morphology, two complementary geometry-based point cloud metrics are introduced: local point density, which highlights poor surface finish, and normal-change rate, which identifies lack-of-fusion features. These methods enable automated, annotation-free identification of common deposition anomalies directly from reconstructed surfaces, without the need for manual labeling. By directly linking geometric deviation to defect formation, the approach enables precise anomaly localization and advances the feasibility of closed-loop process control. This work establishes fringe projection as a practical tool for micrometer-scale monitoring in DED, bridging the gap between process signatures and part geometry for certifiable additive manufacturing.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 02:26:18 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Hu",
"Guanzhong",
""
],
[
"Li",
"Wenpan",
""
],
[
"Zha",
"Rujing",
""
],
[
"Guo",
"Ping",
""
]
] |
2509.05328
|
Xiang Yuan
|
Xiang Yuan, Jun Shu, Deyu meng, Zongben Xu
|
Feed Two Birds with One Scone: Exploiting Function-Space Regularization for Both OOD Robustness and ID Fine-Tuning Performance
| null | null | null | null |
cs.LG cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Robust fine-tuning aims to achieve competitive in-distribution (ID) performance while maintaining the out-of-distribution (OOD) robustness of a pre-trained model when transferring it to a downstream task. To remedy this, most robust fine-tuning methods aim to preserve the pretrained weights, features, or logits. However, we find that these methods cannot always improve OOD robustness for different model architectures. This is due to the OOD robustness requiring the model function to produce stable prediction for input information of downstream tasks, while existing methods might serve as a poor proxy for the optimization in the function space. Based on this finding, we propose a novel regularization that constrains the distance of fine-tuning and pre-trained model in the function space with the simulated OOD samples, aiming to preserve the OOD robustness of the pre-trained model. Besides, to further enhance the OOD robustness capability of the fine-tuning model, we introduce an additional consistency regularization to promote stable predictions of perturbed samples. Extensive experiments demonstrate our approach could consistently improve both downstream task ID fine-tuning performance and OOD robustness across a variety of CLIP backbones, outperforming existing regularization-based robust fine-tuning methods.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 12:14:34 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Yuan",
"Xiang",
""
],
[
"Shu",
"Jun",
""
],
[
"meng",
"Deyu",
""
],
[
"Xu",
"Zongben",
""
]
] |
2509.05329
|
Francesco Foscarin
|
Juan Carlos Martinez-Sevilla, Francesco Foscarin, Patricia Garcia-Iasci, David Rizo, Jorge Calvo-Zaragoza, Gerhard Widmer
|
Optical Music Recognition of Jazz Lead Sheets
|
Accepted at the 26th International Society for Music Information Retrieval Conference (ISMIR), 2025
| null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we address the challenge of Optical Music Recognition (OMR) for handwritten jazz lead sheets, a widely used musical score type that encodes melody and chords. The task is challenging due to the presence of chords, a score component not handled by existing OMR systems, and the high variability and quality issues associated with handwritten images. Our contribution is two-fold. We present a novel dataset consisting of 293 handwritten jazz lead sheets of 163 unique pieces, amounting to 2021 total staves aligned with Humdrum **kern and MusicXML ground truth scores. We also supply synthetic score images generated from the ground truth. The second contribution is the development of an OMR model for jazz lead sheets. We discuss specific tokenisation choices related to our kind of data, and the advantages of using synthetic scores and pretrained models. We publicly release all code, data, and models.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 14:38:26 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Martinez-Sevilla",
"Juan Carlos",
""
],
[
"Foscarin",
"Francesco",
""
],
[
"Garcia-Iasci",
"Patricia",
""
],
[
"Rizo",
"David",
""
],
[
"Calvo-Zaragoza",
"Jorge",
""
],
[
"Widmer",
"Gerhard",
""
]
] |
2509.05330
|
S Muhammad Hossein Mousavi
|
Seyed Muhammad Hossein Mousavi, Atiye Ilanloo
|
MVRS: The Multimodal Virtual Reality Stimuli-based Emotion Recognition Dataset
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Automatic emotion recognition has become increasingly important with the rise of AI, especially in fields like healthcare, education, and automotive systems. However, there is a lack of multimodal datasets, particularly involving body motion and physiological signals, which limits progress in the field. To address this, the MVRS dataset is introduced, featuring synchronized recordings from 13 participants aged 12 to 60 exposed to VR based emotional stimuli (relaxation, fear, stress, sadness, joy). Data were collected using eye tracking (via webcam in a VR headset), body motion (Kinect v2), and EMG and GSR signals (Arduino UNO), all timestamp aligned. Participants followed a unified protocol with consent and questionnaires. Features from each modality were extracted, fused using early and late fusion techniques, and evaluated with classifiers to confirm the datasets quality and emotion separability, making MVRS a valuable contribution to multimodal affective computing.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 17:20:53 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Mousavi",
"Seyed Muhammad Hossein",
""
],
[
"Ilanloo",
"Atiye",
""
]
] |
2509.05331
|
Youssef Chakir
|
Youssef Chakir and Iyad Lahsen-Cherif
|
ForensicsData: A Digital Forensics Dataset for Large Language Models
|
Accepted to WiMob 2025 (21st International Conference on Wireless and Mobile Computing, Networking and Communications), Marrakesh, Morocco, Oct 20-22, 2025. 6 pages, 5 figures, 5 tables. IEEEtran conference format
| null | null | null |
cs.CR cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
The growing complexity of cyber incidents presents significant challenges for digital forensic investigators, especially in evidence collection and analysis. Public resources are still limited because of ethical, legal, and privacy concerns, even though realistic datasets are necessary to support research and tool developments. To address this gap, we introduce ForensicsData, an extensive Question-Context-Answer (Q-C-A) dataset sourced from actual malware analysis reports. It consists of more than 5,000 Q-C-A triplets. A unique workflow was used to create the dataset, which extracts structured data, uses large language models (LLMs) to transform it into Q-C-A format, and then uses a specialized evaluation process to confirm its quality. Among the models evaluated, Gemini 2 Flash demonstrated the best performance in aligning generated content with forensic terminology. ForensicsData aims to advance digital forensics by enabling reproducible experiments and fostering collaboration within the research community.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 19:58:24 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Chakir",
"Youssef",
""
],
[
"Lahsen-Cherif",
"Iyad",
""
]
] |
2509.05332
|
Christos Anagnostopoulos
|
Christos Anagnostopoulos, Ioulia Kapsali, Alexandros Gkillas, Nikos Piperigkos, Aris S. Lalos
|
Integrated Simulation Framework for Adversarial Attacks on Autonomous Vehicles
|
6 pages, 2 figures
| null | null | null |
cs.CR cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Autonomous vehicles (AVs) rely on complex perception and communication systems, making them vulnerable to adversarial attacks that can compromise safety. While simulation offers a scalable and safe environment for robustness testing, existing frameworks typically lack comprehensive supportfor modeling multi-domain adversarial scenarios. This paper introduces a novel, open-source integrated simulation framework designed to generate adversarial attacks targeting both perception and communication layers of AVs. The framework provides high-fidelity modeling of physical environments, traffic dynamics, and V2X networking, orchestrating these components through a unified core that synchronizes multiple simulators based on a single configuration file. Our implementation supports diverse perception-level attacks on LiDAR sensor data, along with communication-level threats such as V2X message manipulation and GPS spoofing. Furthermore, ROS 2 integration ensures seamless compatibility with third-party AV software stacks. We demonstrate the framework's effectiveness by evaluating the impact of generated adversarial scenarios on a state-of-the-art 3D object detector, revealing significant performance degradation under realistic conditions.
|
[
{
"version": "v1",
"created": "Sun, 31 Aug 2025 20:53:08 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Anagnostopoulos",
"Christos",
""
],
[
"Kapsali",
"Ioulia",
""
],
[
"Gkillas",
"Alexandros",
""
],
[
"Piperigkos",
"Nikos",
""
],
[
"Lalos",
"Aris S.",
""
]
] |
2509.05333
|
Tuan Anh Nguyen
|
Junghyun Park, Tuan Anh Nguyen, Dugki Min
|
RT-VLM: Re-Thinking Vision Language Model with 4-Clues for Real-World Object Recognition Robustness
| null | null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Real world deployments often expose modern object recognition models to domain shifts that precipitate a severe drop in accuracy. Such shifts encompass (i) variations in low level image statistics, (ii) changes in object pose and viewpoint, (iii) partial occlusion, and (iv) visual confusion across adjacent classes. To mitigate this degradation, we introduce the Re-Thinking Vision Language Model (RT-VLM) framework. The foundation of this framework is a unique synthetic dataset generation pipeline that produces images annotated with "4-Clues": precise bounding boxes, class names, detailed object-level captions, and a comprehensive context-level caption for the entire scene. We then perform parameter efficient supervised tuning of Llama 3.2 11B Vision Instruct on this resource. At inference time, a two stage Re-Thinking scheme is executed: the model first emits its own four clues, then re examines these responses as evidence and iteratively corrects them. Across robustness benchmarks that isolate individual domain shifts, RT-VLM consistently surpasses strong baselines. These findings indicate that the integration of structured multimodal evidence with an explicit self critique loop constitutes a promising route toward reliable and transferable visual understanding.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 02:13:00 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Park",
"Junghyun",
""
],
[
"Nguyen",
"Tuan Anh",
""
],
[
"Min",
"Dugki",
""
]
] |
2509.05334
|
Diwen Huang
|
Diwen Huang
|
A Real-Time, Vision-Based System for Badminton Smash Speed Estimation on Mobile Devices
|
6 pages, 3 figures, 1 table. Independent research preprint
| null | null | null |
cs.CV cs.MM
|
http://creativecommons.org/licenses/by/4.0/
|
Performance metrics in sports, such as shot speed and angle, provide crucial feedback for athlete development. However, the technology to capture these metrics has historically been expensive, complex, and largely inaccessible to amateur and recreational players. This paper addresses this gap in the context of badminton, one of the world's most popular sports, by introducing a novel, cost-effective, and user-friendly system for measuring smash speed using ubiquitous smartphone technology. Our approach leverages a custom-trained YOLOv5 model for shuttlecock detection, combined with a Kalman filter for robust trajectory tracking. By implementing a video-based kinematic speed estimation method with spatiotemporal scaling, the system automatically calculates the shuttlecock's velocity from a standard video recording. The entire process is packaged into an intuitive mobile application, democratizing access to high-level performance analytics and empowering players at all levels to analyze and improve their game.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 06:09:19 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Huang",
"Diwen",
""
]
] |
2509.05335
|
Zebo Xu
|
Zebo Xu, Shaoyun Yu, Mark Torrance, Guido Nottbusch, Nan Zhao, and Zhenguang Cai
|
A Stroke-Level Large-Scale Database of Chinese Character Handwriting and the OpenHandWrite_Toolbox for Handwriting Research
| null | null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Understanding what linguistic components (e.g., phonological, semantic, and orthographic systems) modulate Chinese handwriting at the character, radical, and stroke levels remains an important yet understudied topic. Additionally, there is a lack of comprehensive tools for capturing and batch-processing fine-grained handwriting data. To address these issues, we constructed a large-scale handwriting database in which 42 Chinese speakers for each handwriting 1200 characters in a handwriting-to-dictation task. Additionally, we enhanced the existing handwriting package and provided comprehensive documentation for the upgraded OpenHandWrite_Toolbox, which can easily modify the experimental design, capture the stroke-level handwriting trajectory, and batch-process handwriting measurements (e.g., latency, duration, and pen-pressure). In analysing our large-scale database, multiple regression results show that orthographic predictors impact handwriting preparation and execution across character, radical, and stroke levels. Phonological factors also influence execution at all three levels. Importantly, these lexical effects demonstrate hierarchical attenuation - they were most pronounced at the character level, followed by the radical, and were weakest at the stroke levels. These findings demonstrate that handwriting preparation and execution at the radical and stroke levels are closely intertwined with linguistic components. This database and toolbox offer valuable resources for future psycholinguistic and neurolinguistic research on the handwriting of characters and sub-characters across different languages.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 07:19:37 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Xu",
"Zebo",
""
],
[
"Yu",
"Shaoyun",
""
],
[
"Torrance",
"Mark",
""
],
[
"Nottbusch",
"Guido",
""
],
[
"Zhao",
"Nan",
""
],
[
"Cai",
"Zhenguang",
""
]
] |
2509.05336
|
Gustav Graeber
|
Johannes Baller, Andr\'e Hilger, Naiyu Qi, Chiara Morini, Andrea Cornelio, Arndt Remhof, Markus Osenberg, Ingo Manke, Julian Moosmann, Felix Beckmann, Gustav Graeber
|
Wetting Interactions Between Porous Carbon Hosts and Liquid Sodium-Potassium Alloys Toward Their Use in Negative Electrodes of Alkali-Metal Batteries
| null | null | null | null |
cond-mat.soft cond-mat.mtrl-sci
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Batteries with liquid alkali-metal negative electrodes offer a route to compact, high-performance energy storage. Innovation in alkali-metal management, i.e., controlled storage, release and transport of liquid alkali metal, can enable simpler and cheaper cell designs. Porous carbons have emerged as potential host materials for liquid alkali metals. Here, we study the wetting interactions between porous carbon hosts and liquid sodium-potassium alloy (NaK) as a function of carbon host morphology and surface functionalization via X-ray computed tomography. While as-received carbon samples show no affinity towards NaK, heat-treated carbon is spontaneously infiltrated with NaK filling almost the entire pore volume. We explore how forced wetting partially fills pores of NaK-repellent hosts, showing large differences in pore filling based on the average pore size of the host material. In electrochemical discharge experiments, we show that both as-received and heat-treated carbon felt enable high areal capacities beyond 40 mAh cm-2. However, the heat-treated carbon shows ten times lower overpotential. Finally, we demonstrate how heat-treated carbon felt can enable capillary transport of NaK. In summary, this study elucidates important aspects of the interactions between liquid alkali metals and porous carbon hosts, generating insights into possible applications in liquid alkali-metal batteries.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 12:47:17 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Baller",
"Johannes",
""
],
[
"Hilger",
"André",
""
],
[
"Qi",
"Naiyu",
""
],
[
"Morini",
"Chiara",
""
],
[
"Cornelio",
"Andrea",
""
],
[
"Remhof",
"Arndt",
""
],
[
"Osenberg",
"Markus",
""
],
[
"Manke",
"Ingo",
""
],
[
"Moosmann",
"Julian",
""
],
[
"Beckmann",
"Felix",
""
],
[
"Graeber",
"Gustav",
""
]
] |
2509.05337
|
Gokhan Solak
|
Younggeol Cho, Gokhan Solak, Olivia Nocentini, Marta Lorenzini, Andrea Fortuna, Arash Ajoudani
|
Anticipatory Fall Detection in Humans with Hybrid Directed Graph Neural Networks and Long Short-Term Memory
|
Presented at IEEE RO-MAN 2025
| null | null | null |
cs.CV cs.RO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Detecting and preventing falls in humans is a critical component of assistive robotic systems. While significant progress has been made in detecting falls, the prediction of falls before they happen, and analysis of the transient state between stability and an impending fall remain unexplored. In this paper, we propose a anticipatory fall detection method that utilizes a hybrid model combining Dynamic Graph Neural Networks (DGNN) with Long Short-Term Memory (LSTM) networks that decoupled the motion prediction and gait classification tasks to anticipate falls with high accuracy. Our approach employs real-time skeletal features extracted from video sequences as input for the proposed model. The DGNN acts as a classifier, distinguishing between three gait states: stable, transient, and fall. The LSTM-based network then predicts human movement in subsequent time steps, enabling early detection of falls. The proposed model was trained and validated using the OUMVLP-Pose and URFD datasets, demonstrating superior performance in terms of prediction error and recognition accuracy compared to models relying solely on DGNN and models from literature. The results indicate that decoupling prediction and classification improves performance compared to addressing the unified problem using only the DGNN. Furthermore, our method allows for the monitoring of the transient state, offering valuable insights that could enhance the functionality of advanced assistance systems.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 12:56:31 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Cho",
"Younggeol",
""
],
[
"Solak",
"Gokhan",
""
],
[
"Nocentini",
"Olivia",
""
],
[
"Lorenzini",
"Marta",
""
],
[
"Fortuna",
"Andrea",
""
],
[
"Ajoudani",
"Arash",
""
]
] |
2509.05338
|
Atsushi Masumori
|
Atsushi Masumori, Norihiro Maruyama, Itsuki Doi, johnsmith, Hiroki Sato, Takashi Ikegami
|
Plantbot: Integrating Plant and Robot through LLM Modular Agent Networks
| null | null | null | null |
cs.RO cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We introduce Plantbot, a hybrid lifeform that connects a living plant with a mobile robot through a network of large language model (LLM) modules. Each module - responsible for sensing, vision, dialogue, or action - operates asynchronously and communicates via natural language, enabling seamless interaction across biological and artificial domains. This architecture leverages the capacity of LLMs to serve as hybrid interfaces, where natural language functions as a universal protocol, translating multimodal data (soil moisture, temperature, visual context) into linguistic messages that coordinate system behaviors. The integrated network transforms plant states into robotic actions, installing normativity essential for agency within the sensor-motor loop. By combining biological and robotic elements through LLM-mediated communication, Plantbot behaves as an embodied, adaptive agent capable of responding autonomously to environmental conditions. This approach suggests possibilities for a new model of artificial life, where decentralized, LLM modules coordination enable novel interactions between biological and artificial systems.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 14:12:28 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Masumori",
"Atsushi",
""
],
[
"Maruyama",
"Norihiro",
""
],
[
"Doi",
"Itsuki",
""
],
[
"johnsmith",
"",
""
],
[
"Sato",
"Hiroki",
""
],
[
"Ikegami",
"Takashi",
""
]
] |
2509.05339
|
Marco Salvalaglio
|
Abel H. G. Milor, Otto Sumray, Heather A. Harrington, Axel Voigt, Marco Salvalaglio
|
A topological approach to the Cahn-Hilliard equation and hyperuniform fields
|
20 pages, 6 figures
| null | null | null |
cond-mat.soft cond-mat.dis-nn cond-mat.stat-mech math.AP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Hyperuniform structures are disordered, correlated systems in which density fluctuations are suppressed at large scales. Such a property generalizes the concept of order in patterns and is relevant across diverse physical systems. We present a numerical characterization of hyperuniform scalar fields that leverages persistent homology. Topological features across different lengths are represented in persistence diagrams, while similarities or differences between patterns are quantified through Wasserstein distances between these diagrams. We apply this framework to numerical solutions of the Cahn-Hilliard equation, a canonical model for generating hyperuniform scalar fields. We validate the approach against known features of the Cahn-Hilliard equation, including its scaling properties, convergence to the sharp interface limit, and self-similarity of the solutions. We then generalize the approach by studying Gaussian random fields exhibiting different degrees and classes of hyperuniformity, showing how the proposed approach can be exploited to reconstruct global properties from local topological information. Overall, we show how hyperuniform characteristics systematically correlate with distributions of topological features in disordered correlated fields. We expect this analysis to be applicable to a wide range of scalar fields, particularly those involving interfaces and free boundaries.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 14:42:29 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Milor",
"Abel H. G.",
""
],
[
"Sumray",
"Otto",
""
],
[
"Harrington",
"Heather A.",
""
],
[
"Voigt",
"Axel",
""
],
[
"Salvalaglio",
"Marco",
""
]
] |
2509.05340
|
Dibya Jyoti Bora
|
Dibya Jyoti Bora, Mrinal Kanti Mishra
|
Comparative Evaluation of Hard and Soft Clustering for Precise Brain Tumor Segmentation in MR Imaging
|
15 pages, 10 figures
|
Journal of Advances in Mathematics and Computer Science 40 (9) 2025
|
10.9734/jamcs/2025/v40i92050
| null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Segmentation of brain tumors from Magnetic Resonance Imaging (MRI) remains a pivotal challenge in medical image analysis due to the heterogeneous nature of tumor morphology and intensity distributions. Accurate delineation of tumor boundaries is critical for clinical decision-making, radiotherapy planning, and longitudinal disease monitoring. In this study, we perform a comprehensive comparative analysis of two major clustering paradigms applied in MRI tumor segmentation: hard clustering, exemplified by the K-Means algorithm, and soft clustering, represented by Fuzzy C-Means (FCM). While K-Means assigns each pixel strictly to a single cluster, FCM introduces partial memberships, meaning each pixel can belong to multiple clusters with varying degrees of association. Experimental validation was performed using the BraTS2020 dataset, incorporating pre-processing through Gaussian filtering and Contrast Limited Adaptive Histogram Equalization (CLAHE). Evaluation metrics included the Dice Similarity Coefficient (DSC) and processing time, which collectively demonstrated that K-Means achieved superior speed with an average runtime of 0.3s per image, whereas FCM attained higher segmentation accuracy with an average DSC of 0.67 compared to 0.43 for K-Means, albeit at a higher computational cost (1.3s per image). These results highlight the inherent trade-off between computational efficiency and boundary precision.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 16:40:08 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Bora",
"Dibya Jyoti",
""
],
[
"Mishra",
"Mrinal Kanti",
""
]
] |
2509.05341
|
Abhijeet Pal
|
Abhijeet Manoj Pal, Rajbabu Velmurugan
|
Handling imbalance and few-sample size in ML based Onion disease classification
|
6 pages, 8 figures
| null | null | null |
cs.CV cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Accurate classification of pests and diseases plays a vital role in precision agriculture, enabling efficient identification, targeted interventions, and preventing their further spread. However, current methods primarily focus on binary classification, which limits their practical applications, especially in scenarios where accurately identifying the specific type of disease or pest is essential. We propose a robust deep learning based model for multi-class classification of onion crop diseases and pests. We enhance a pre-trained Convolutional Neural Network (CNN) model by integrating attention based modules and employing comprehensive data augmentation pipeline to mitigate class imbalance. We propose a model which gives 96.90% overall accuracy and 0.96 F1 score on real-world field image dataset. This model gives better results than other approaches using the same datasets.
|
[
{
"version": "v1",
"created": "Mon, 01 Sep 2025 19:05:39 GMT"
}
] | 2025-09-09T00:00:00 |
[
[
"Pal",
"Abhijeet Manoj",
""
],
[
"Velmurugan",
"Rajbabu",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.