id
stringlengths 9
16
| submitter
stringlengths 1
64
⌀ | authors
stringlengths 3
62.8k
| title
stringlengths 1
382
| comments
stringlengths 1
1.15k
⌀ | journal-ref
stringlengths 1
557
⌀ | doi
stringlengths 7
153
⌀ | report-no
stringlengths 1
509
⌀ | categories
stringlengths 5
125
| license
stringclasses 9
values | abstract
stringlengths 3
6.09k
| versions
listlengths 1
187
| update_date
timestamp[s]date 2007-05-23 00:00:00
2025-08-22 00:00:00
| authors_parsed
listlengths 1
3.31k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2508.14226
|
Daniel Baker
|
Daniel Baker (1), Jeremy Wojcik (1), Sean Phillips (2) ((1) Defense Sector, BlueHalo, LLC, (2) Space Control Branch (RVSW), Air Force Research Laboratory)
|
Autonomy at Levels for Spacecraft
|
18 pages, 10 figures, to be published in 2025 AAS/AIAA Astrodynamics Specialist conference proceedings
| null | null |
AFRL-2025-3557
|
eess.SY cs.SY
|
http://creativecommons.org/licenses/by/4.0/
|
Autonomy at Levels is the idea that autonomy should be embedded within and throughout a spacecraft. Using Systems Engineering methods a spacecraft is typically decomposed into systems, subsystems, assemblies, components, and so on. All these decomposition levels within all the spacecraft's systems, could and should have autonomy elements built in. As a result, the "autonomy system" is made of autonomy elements or units that are integrated, distributed and embedded within the whole spacecraft. This is like how the power system would be designed and implemented. Linking control loops and autonomy loops illustrates how to achieve Autonomy at Levels.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:33:57 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Baker",
"Daniel",
"",
"Defense Sector, BlueHalo, LLC"
],
[
"Wojcik",
"Jeremy",
"",
"Defense Sector, BlueHalo, LLC"
],
[
"Phillips",
"Sean",
"",
"Space Control Branch"
]
] |
2508.14227
|
Tyler LeBlond
|
Tyler LeBlond, Peter Groszkowski, Justin G. Lietz, Christopher M. Seck, Ryan S. Bennink
|
Logical Error Rates for the Surface Code Under a Mixed Coherent and Stochastic Circuit-Level Noise Model Inspired by Trapped Ions
|
20 pages, 8 figures
| null | null | null |
quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
With fault-tolerant quantum computing (FTQC) on the horizon, it is critical to understand sources of logical error in plausible hardware implementations of quantum error-correcting codes (QECC). In this work, we consider logical error rates for the surface code implemented on a hypothetical grid-based trapped-ion quantum charge-coupled device (QCCD) architecture. Specifically, we construct logical channels for the idling surface code and examine its diamond error under a mixed coherent and stochastic circuit-level noise model inspired by trapped ions. We include the coherent dephasing noise that is known to accumulate during physical qubit idling and transport in these systems, determining idling and transport durations using the time-resolved output of the trapped-ion surface code compiler (TISCC). To estimate expectation values of logical Pauli observables following hardware circuits containing non-Clifford sources of noise, we utilize a Monte Carlo technique to sample from an underlying quasi-probability distribution of Clifford circuits that we independently simulate in a phase-sensitive fashion. We verify error suppression up to code distance $d=11$ at coherent dephasing rates near and below those of current-generation trapped-ion quantum computers and find that logical error rates align with those of analogous fully stochastic simulations in this regime. Exploring higher dephasing rates at $d=3-5$, we find evidence for growing coherent rotations about all three logical Pauli axes, increased diagonal logical error process matrix elements relative to those of stochastic simulations, and a reduced dephasing rate threshold. Overall, our work paves a way toward realistic hardware emulation of small fault-tolerant quantum processes, e.g., members of a FTQC instruction set.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:37:09 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"LeBlond",
"Tyler",
""
],
[
"Groszkowski",
"Peter",
""
],
[
"Lietz",
"Justin G.",
""
],
[
"Seck",
"Christopher M.",
""
],
[
"Bennink",
"Ryan S.",
""
]
] |
2508.14228
|
Soham Bhattacharyya
|
Soham Bhattacharyya, Soumitra SenGupta
|
Thermal description of braneworld effective theories
|
10 pages, 3 figures, feedback welcome
| null | null | null |
hep-th gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Low-energy effective theories provide the natural description of four-dimensional physics in higher-dimensional geometries, where the imprint of the bulk appears as parameters of the lower dimensional theory. Motivated by the recent progress in the first-order thermodynamic formulation of modified gravity theories, we investigate the thermodynamics of effective theories in braneworld scenarios and thereby the attractor mechanism towards general relativity in such theories. We consider the two-brane Randall-Sundrum model where the low-energy theory on either brane is of scalar-tensor nature with the extra-dimensional radion playing the role of the scalar. We study the thermodynamic implications of a non-vanishing gravitational contribution to the radion potential, and further explore the dynamics in the presence of a bulk stabilizing field.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:37:53 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Bhattacharyya",
"Soham",
""
],
[
"SenGupta",
"Soumitra",
""
]
] |
2508.14229
|
Md Mainul Abrar
|
Md Mainul Abrar, Xun Jia, Yujie Chi
|
New Insights into Automatic Treatment Planning for Cancer Radiotherapy Using Explainable Artificial Intelligence
|
19 pages, 7 figures, 1 table, Oral presentation at the conference 'American Association of Physicists in Medicine 2025, 67th Annual Meeting and Exhibition'
| null | null | null |
physics.med-ph cs.AI
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Objective: This study aims to uncover the opaque decision-making process of an artificial intelligence (AI) agent for automatic treatment planning.
Approach: We examined a previously developed AI agent based on the Actor-Critic with Experience Replay (ACER) network, which automatically tunes treatment planning parameters (TPPs) for inverse planning in prostate cancer intensity modulated radiotherapy. We selected multiple checkpoint ACER agents from different stages of training and applied an explainable AI (EXAI) method to analyze the attribution from dose-volume histogram (DVH) inputs to TPP-tuning decisions. We then assessed each agent's planning efficacy and efficiency and evaluated their policy and final TPP tuning spaces. Combining these analyses, we systematically examined how ACER agents generated high-quality treatment plans in response to different DVH inputs.
Results: Attribution analysis revealed that ACER agents progressively learned to identify dose-violation regions from DVH inputs and promote appropriate TPP-tuning actions to mitigate them. Organ-wise similarities between DVH attributions and dose-violation reductions ranged from 0.25 to 0.5 across tested agents. Agents with stronger attribution-violation similarity required fewer tuning steps (~12-13 vs. 22), exhibited a more concentrated TPP-tuning space with lower entropy (~0.3 vs. 0.6), converged on adjusting only a few TPPs, and showed smaller discrepancies between practical and theoretical tuning steps. Putting together, these findings indicate that high-performing ACER agents can effectively identify dose violations from DVH inputs and employ a global tuning strategy to achieve high-quality treatment planning, much like skilled human planners.
Significance: Better interpretability of the agent's decision-making process may enhance clinician trust and inspire new strategies for automatic treatment planning.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:38:16 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Abrar",
"Md Mainul",
""
],
[
"Jia",
"Xun",
""
],
[
"Chi",
"Yujie",
""
]
] |
2508.14230
|
Eduardo Ribas Brito
|
Eduardo Brito, Fernando Castillo, Liina Kamm, Amnir Hadachi, Ulrich Norbisrath
|
A Taxonomy and Methodology for Proof-of-Location Systems
|
This work has been accepted to the 29th International Conference on Enterprise Design, Operations, and Computing (EDOC 2025)
| null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Digital societies increasingly rely on trustworthy proofs of physical presence for services such as supply-chain tracking, e-voting, ride-sharing, and location-based rewards. Yet, traditional localization methods often lack cryptographic guarantees of where and when an entity was present, leaving them vulnerable to spoofing, replay, or collusion attacks. In response, research on Proof-of-Location (PoL) has emerged, with recent approaches combining distance bounding, distributed consensus, and privacy-enhancing techniques to enable verifiable, tamper-resistant location claims.
As the design space for PoL systems grows in complexity, this paper provides a unified framework to help practitioners navigate diverse application needs. We first propose a taxonomy identifying four core domains: (1) cryptographic guarantees, (2) spatio-temporal synchronization, (3) trust and witness models, and (4) interaction and overhead. Building on this, we introduce a methodology to map application-specific requirements onto appropriate PoL architectures. We illustrate this process through three use cases (retail e-coupons, supply chain auditing, and physical e-voting), each showing how different constraints shape protocol choices. Overall, this work offers a structured approach to building secure, scalable, and interoperable PoL systems.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:39:05 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Brito",
"Eduardo",
""
],
[
"Castillo",
"Fernando",
""
],
[
"Kamm",
"Liina",
""
],
[
"Hadachi",
"Amnir",
""
],
[
"Norbisrath",
"Ulrich",
""
]
] |
2508.14231
|
Carson Ezell
|
Carson Ezell, Xavier Roberts-Gaal, Alan Chan
|
Incident Analysis for AI Agents
|
16 pages (10 pages main text), 4 figures, 3 tables. To be published in the Proceedings of the 2025 AAAI/ACM Conference on AI, Ethics, & Society (AIES)
| null | null | null |
cs.CY cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
As AI agents become more widely deployed, we are likely to see an increasing number of incidents: events involving AI agent use that directly or indirectly cause harm. For example, agents could be prompt-injected to exfiltrate private information or make unauthorized purchases. Structured information about such incidents (e.g., user prompts) can help us understand their causes and prevent future occurrences. However, existing incident reporting processes are not sufficient for understanding agent incidents. In particular, such processes are largely based on publicly available data, which excludes useful, but potentially sensitive, information such as an agent's chain of thought or browser history. To inform the development of new, emerging incident reporting processes, we propose an incident analysis framework for agents. Drawing on systems safety approaches, our framework proposes three types of factors that can cause incidents: system-related (e.g., CBRN training data), contextual (e.g., prompt injections), and cognitive (e.g., misunderstanding a user request). We also identify specific information that could help clarify which factors are relevant to a given incident: activity logs, system documentation and access, and information about the tools an agent uses. We provide recommendations for 1) what information incident reports should include and 2) what information developers and deployers should retain and make available to incident investigators upon request. As we transition to a world with more agents, understanding agent incidents will become increasingly crucial for managing risks.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:39:37 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Ezell",
"Carson",
""
],
[
"Roberts-Gaal",
"Xavier",
""
],
[
"Chan",
"Alan",
""
]
] |
2508.14232
|
Matthew Andres Moreno
|
Matthew Andres Moreno and Sanaz Hasanzadeh Fard and Luis Zaman and Emily Dolson
|
Extending a Phylogeny-based Method for Detecting Signatures of Multi-level Selection for Applications in Artificial Life
| null | null | null | null |
q-bio.PE cs.NE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Multilevel selection occurs when short-term individual-level reproductive interests conflict with longer-term group-level fitness effects. Detecting and quantifying this phenomenon is key to understanding evolution of traits ranging from multicellularity to pathogen virulence. Multilevel selection is particularly important in artificial life research due to its connection to major evolutionary transitions, a hallmark of open-ended evolution. Bonetti Franceschi & Volz (2024) proposed to detect multilevel selection dynamics by screening for mutations that appear more often in a population than expected by chance (due to individual-level fitness benefits) but are ultimately associated with negative longer-term fitness outcomes (i.e., smaller, shorter-lived descendant clades). Here, we use agent-based modeling with known ground truth to assess the efficacy of this approach. To test these methods under challenging conditions broadly comparable to the original dataset explored by Bonetti Franceschi & Volz (2024), we use an epidemiological framework to model multilevel selection in trade-offs between within-host growth rate and between-host transmissibility. To achieve success on our in silico data, we develop an alternate normalization procedure for identifying clade-level fitness effects. We find the method to be sensitive in detecting genome sites under multilevel selection with 30% effect sizes on fitness, but do not see sensitivity to smaller 10% mutation effect sizes. To test the robustness of this methodology, we conduct additional experiments incorporating extrinsic, time-varying environmental changes and adaptive turnover in population compositions, and find that screen performance remains generally consistent with baseline conditions. This work represents a promising step towards rigorous generalizable quantification of multilevel selection effects.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:42:38 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Moreno",
"Matthew Andres",
""
],
[
"Fard",
"Sanaz Hasanzadeh",
""
],
[
"Zaman",
"Luis",
""
],
[
"Dolson",
"Emily",
""
]
] |
2508.14233
|
Ian Abrahams
|
Ian T. Abrahams
|
Excitonic Coupling and Photon Antibunching in Venus Yellow Fluorescent Protein Dimers: A Lindblad Master Equation Approach
|
12 pages (excluding references), 7 figures, includes discussions of cryogenic exciton dynamics, quantum biophotonics, quantum technology, evolutionary adaptations in fluorescent proteins, and the potential application of Venus dimers as quantum bits (qubits) for quantum information processing
| null | null | null |
quant-ph cond-mat.mes-hall physics.bio-ph physics.optics q-bio.BM
|
http://creativecommons.org/licenses/by/4.0/
|
Strong excitonic coupling and photon antibunching have been observed together in Venus yellow fluorescent protein (YFP) dimers -- a cryptic combination under prevailing theoretical models. In 2019, Kim et al. demonstrated Davydov splitting in Venus dimer circular dichroism (CD) spectra, revealing large negative dimer coupling energy, while antibunching was confirmed by antibunching-fluorescence correlation spectroscopy (AB/FCS fingerprinting). To explain this coexistence, Venus dimer population dynamics are modeled here within a Lindblad master equation framework, justified by the separation of characteristic coupling, dephasing, and thermal relaxation rates. Simulations predict rapid decoherence, consistent with antibunching, indicating that coherence in both the excitonic and site bases is too short-lived to affect photon emission statistics. Despite the absence of long-lived coherence, Venus dimers provide a tractable platform for probing evolutionary pressures on fluorescent protein photophysics and quantum dynamics. Cryogenic cooling could extend coherence lifetimes into the regime required for quantum gate operations, suggesting a route toward fluorescent protein-based qubits. More broadly, the results highlight a structural design principle -- a "bioexciton motif" -- that links excitonic coupling, decoherence, and protein architecture, pointing to general rules by which biology both constrains and inspires quantum technologies.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:44:59 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Abrahams",
"Ian T.",
""
]
] |
2508.14234
|
Shabarish Chenakkod
|
Shabarish Chenakkod, Micha{\l} Derezi\'nski, Xiaoyu Dong
|
Optimal Subspace Embeddings: Resolving Nelson-Nguyen Conjecture Up to Sub-Polylogarithmic Factors
| null | null | null | null |
cs.DS cs.LG cs.NA math.NA math.PR stat.ML
|
http://creativecommons.org/licenses/by/4.0/
|
We give a proof of the conjecture of Nelson and Nguyen [FOCS 2013] on the optimal dimension and sparsity of oblivious subspace embeddings, up to sub-polylogarithmic factors: For any $n\geq d$ and $\epsilon\geq d^{-O(1)}$, there is a random $\tilde O(d/\epsilon^2)\times n$ matrix $\Pi$ with $\tilde O(\log(d)/\epsilon)$ non-zeros per column such that for any $A\in\mathbb{R}^{n\times d}$, with high probability, $(1-\epsilon)\|Ax\|\leq\|\Pi Ax\|\leq(1+\epsilon)\|Ax\|$ for all $x\in\mathbb{R}^d$, where $\tilde O(\cdot)$ hides only sub-polylogarithmic factors in $d$. Our result in particular implies a new fastest sub-current matrix multiplication time reduction of size $\tilde O(d/\epsilon^2)$ for a broad class of $n\times d$ linear regression tasks.
A key novelty in our analysis is a matrix concentration technique we call iterative decoupling, which we use to fine-tune the higher-order trace moment bounds attainable via existing random matrix universality tools [Brailovskaya and van Handel, GAFA 2024].
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:45:52 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Chenakkod",
"Shabarish",
""
],
[
"Dereziński",
"Michał",
""
],
[
"Dong",
"Xiaoyu",
""
]
] |
2508.14235
|
Omar Mostafa
|
Omar Mostafa, Nikolaos Evangeliou, Anthony Tzes
|
SLAM-based Safe Indoor Exploration Strategy
|
5 pages, 8 figures. Published in the 2025 11th International Conference on Automation, Robotics, and Applications (ICARA)
|
2025 11th International Conference on Automation, Robotics, and Applications (ICARA), pp. 375-379
|
10.1109/ICARA64554.2025
| null |
cs.RO
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
This paper suggests a 2D exploration strategy for a planar space cluttered with obstacles. Rather than using point robots capable of adjusting their position and altitude instantly, this research is tailored to classical agents with circular footprints that cannot control instantly their pose. Inhere, a self-balanced dual-wheeled differential drive system is used to explore the place. The system is equipped with linear accelerometers and angular gyroscopes, a 3D-LiDAR, and a forward-facing RGB-D camera. The system performs RTAB-SLAM using the IMU and the LiDAR, while the camera is used for loop closures. The mobile agent explores the planar space using a safe skeleton approach that places the agent as far as possible from the static obstacles. During the exploration strategy, the heading is towards any offered openings of the space. This space exploration strategy has as its highest priority the agent's safety in avoiding the obstacles followed by the exploration of undetected space. Experimental studies with a ROS-enabled mobile agent are presented indicating the path planning strategy while exploring the space.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:50:24 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Mostafa",
"Omar",
""
],
[
"Evangeliou",
"Nikolaos",
""
],
[
"Tzes",
"Anthony",
""
]
] |
2508.14236
|
Minyi Huang
|
Minyi Huang, Shuenn-Jyi Sheu, and Li-Hsien Sun
|
Mean field social optimization: feedback person-by-person optimality and the dynamic programming equation
|
The first version appeared as a GERAD Technical Report G-2024-45, Aug 2024 (https://www.gerad.ca); this version includes several minor editorial changes
| null | null | null |
math.OC
|
http://creativecommons.org/licenses/by/4.0/
|
We consider mean field social optimization in nonlinear diffusion models. By dynamic programming with a representative agent employing cooperative optimizer selection, we derive a new Hamilton--Jacobi--Bellman (HJB) equation to be called the master equation of the value function. Under some regularity conditions, we establish $\epsilon$-person-by-person optimality of the master equation-based control laws, which may be viewed as a necessary condition for nearly attaining the social optimum. A major challenge in the analysis is to obtain tight estimates, within an error of $O(1/N)$, of the social cost having order $O(N)$. This will be accomplished by multi-scale analysis via constructing two auxiliary master equations. We illustrate explicit solutions of the master equations for the linear-quadratic (LQ) case, and give an application to systemic risk.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:54:28 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Huang",
"Minyi",
""
],
[
"Sheu",
"Shuenn-Jyi",
""
],
[
"Sun",
"Li-Hsien",
""
]
] |
2508.14237
|
Miao Zhang
|
Miao Zhang, Yifei Zhu, Linfeng Shen, Fangxin Wang, and Jiangchuan Liu
|
OmniSense: Towards Edge-Assisted Online Analytics for 360-Degree Videos
|
10 pages; Accepted by INFOCOM'23
| null | null | null |
cs.NI cs.CV cs.MM eess.IV
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
With the reduced hardware costs of omnidirectional cameras and the proliferation of various extended reality applications, more and more $360^\circ$ videos are being captured. To fully unleash their potential, advanced video analytics is expected to extract actionable insights and situational knowledge without blind spots from the videos. In this paper, we present OmniSense, a novel edge-assisted framework for online immersive video analytics. OmniSense achieves both low latency and high accuracy, combating the significant computation and network resource challenges of analyzing $360^\circ$ videos. Motivated by our measurement insights into $360^\circ$ videos, OmniSense introduces a lightweight spherical region of interest (SRoI) prediction algorithm to prune redundant information in $360^\circ$ frames. Incorporating the video content and network dynamics, it then smartly scales vision models to analyze the predicted SRoIs with optimized resource utilization. We implement a prototype of OmniSense with commodity devices and evaluate it on diverse real-world collected $360^\circ$ videos. Extensive evaluation results show that compared to resource-agnostic baselines, it improves the accuracy by $19.8\%$ -- $114.6\%$ with similar end-to-end latencies. Meanwhile, it hits $2.0\times$ -- $2.4\times$ speedups while keeping the accuracy on par with the highest accuracy of baselines.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:54:40 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zhang",
"Miao",
""
],
[
"Zhu",
"Yifei",
""
],
[
"Shen",
"Linfeng",
""
],
[
"Wang",
"Fangxin",
""
],
[
"Liu",
"Jiangchuan",
""
]
] |
2508.14238
|
Jasem Hamoud Mohamad
|
Jasem Hamoud
|
The Exponential Hyper-Zagreb Indices and Structural Properties of Trees and Bipartite Graphs
|
16 pages, Comments welcome!
| null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
In this paper, we investigate the structural properties of trees and bipartite graphs through the lens of topological indices and combinatorial graph theory. We focus on the First and Second Hyper-Zagreb indices, $HM_1(G)$ and $HM_2(G)$, for trees $T \in T(n, \Delta)$ with $n$ vertices and maximum degree $\Delta$. Key propositions demonstrate that the presence of end-support or support vertices of degree at least three, distinct from a vertex of maximum degree, implies the existence of another tree $T' \in T(n, \Delta)$ with strictly smaller Hyper-Zagreb indices. These results are extended to exponential forms, highlighting the influence of high-degree vertices. Additionally, we explore structural characterizations of trees via degree sequence majorization and $S$-order, establishing conditions for the first and last trees in specific classes. For bipartite graphs, we examine equitable coloring, cycle lengths, and $k$-redundant tree embeddings, supported by theorems on connectivity and minimum degree constraints. The paper also addresses the independence number of bipartite graphs, exterior covers, and competition numbers of complete $r$-partite graphs, providing bounds and structural insights. Finally, we discuss Markov-chain algorithms for generating bipartite graphs and tournaments with prescribed degree sequences, analyzing their mixing times and convergence properties. These results contribute to the understanding of extremal properties and combinatorial structures in graph theory, with applications in chemical graph theory and network analysis.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:55:10 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Hamoud",
"Jasem",
""
]
] |
2508.14239
|
Shengze Wang
|
Shengze Wang, Yi Liu, Xiaoxue Zhang, Liting Hu, and Chen Qian
|
A Distributed Learned Hash Table
| null | null | null | null |
cs.NI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Distributed Hash Tables (DHTs) are pivotal in numerous high-impact key-value applications built on distributed networked systems, offering a decentralized architecture that avoids single points of failure and improves data availability. Despite their widespread utility, DHTs face substantial challenges in handling range queries, which are crucial for applications such as LLM serving, distributed storage, databases, content delivery networks, and blockchains. To address this limitation, we present LEAD, a novel system incorporating learned models within DHT structures to significantly optimize range query performance. LEAD utilizes a recursive machine learning model to map and retrieve data across a distributed system while preserving the inherent order of data. LEAD includes the designs to minimize range query latency and message cost while maintaining high scalability and resilience to network churn. Our comprehensive evaluations, conducted in both testbed implementation and simulations, demonstrate that LEAD achieves tremendous advantages in system efficiency compared to existing range query methods in large-scale distributed systems, reducing query latency and message cost by 80% to 90%+. Furthermore, LEAD exhibits remarkable scalability and robustness against system churn, providing a robust, scalable solution for efficient data retrieval in distributed key-value systems.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:57:34 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Wang",
"Shengze",
""
],
[
"Liu",
"Yi",
""
],
[
"Zhang",
"Xiaoxue",
""
],
[
"Hu",
"Liting",
""
],
[
"Qian",
"Chen",
""
]
] |
2508.14240
|
Andrew Bruce J
|
Andrew James Bruce
|
On a Super-analogue of Carrollian Manifolds
|
13 pages
| null | null | null |
math.DG gr-qc math-ph math.MP
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Inspired by Carrollian geometry, we define super-Carrollian manifolds as a supermanifold with an even degenerate metric such that the kernel is generated by a non-singular odd vector field that is a supersymmetry generator. Alongside other results, we show that compatible affine connections always exist, albeit they must carry torsion.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 19:59:16 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Bruce",
"Andrew James",
""
]
] |
2508.14241
|
Kihong Kim
|
Seulong Kim, Kihong Kim
|
Deterministic time rewinding of waves in time-varying media
|
11 pages, 8 figures
| null | null | null |
physics.optics cond-mat.mes-hall
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Temporal modulation of material parameters offers unprecedented control over wave dynamics, enabling phenomena beyond the capabilities of static systems. Here we introduce and analyze a robust mechanism for time rewinding, whereby a temporally evolved wave is fully restored to its original state through a carefully engineered sequence of temporal modulations. In electromagnetic systems, time rewinding emerges from impedance-matched or anti-matched hierarchical bilayer structures with matched modulation durations, exploiting total transmission or reflection and reversed phase accumulation. In Dirac systems, it arises via complete interband transition driven by time-dependent vector potentials. Unlike time-reversal holography or quantum time mirrors, which produce wave echoes but only partial waveform recovery, our approach achieves deterministic and complete reconstruction of the entire wave state, including both amplitude and phase. Analytical conditions for robust amplitude and phase restoration are derived and validated through simulations of discrete and continuous modulations, demonstrating resilience to modulation complexity and temporal asymmetry. These findings establish a versatile platform for secure information retrieval, temporal cloaking, programmable metamaterials, and wave-based logic devices.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:00:55 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Kim",
"Seulong",
""
],
[
"Kim",
"Kihong",
""
]
] |
2508.14242
|
Seiya Tsukamoto
|
Seiya Tsukamoto, Andrew Toivonen, Holton Griffin, Avyukt Raghuvanshi, Megan Averill, Frank Kerkow, Michael W. Coughlin, Man Leong Chan, Leo Singer
|
Astrophysical or Terrestrial: Machine learning classification of gravitational-wave candidates using multiple-search information
| null | null | null | null |
gr-qc astro-ph.IM
|
http://creativecommons.org/licenses/by/4.0/
|
Low-latency gravitational-wave alerts provide the greater multi-messenger community with information about the candidate events detected by the International Gravitational-Wave Network (IGWN). Prompt release of data products such as the sky localization, false alarm rate (FAR), and $p_\mathrm{astro}$ values allow astronomers to make informed decisions on which candidate gravitational-wave events merit target of opportunity (ToO) follow-up. However, false alarms, often referred to as "glitches", where a gravitational-wave candidate, or trigger, is the result of terrestrial noise, are an inherent part of gravitational-wave searches. In addition, with the presence of multiple gravitational-wave searches, different searches may have varying assessments of the significance of a given trigger. As a complement to quantities such as $p_\mathrm{astro}$, we provide a Machine Learning (ML) based approach to determining whether candidate events are astrophysical or terrestrial in nature, specifically a classifier that utilizes information provided by multiple low-latency search pipelines in its feature space. This classifier has a performance an Area Under the Receiver Operating Characteristic Curve (AUC) of 0.96 and accuracy of 0.90 on the Mock Data Challenge training set and an AUC of 0.93 and accuracy of 0.86 on events from the Advanced LIGO (aLIGO)'s and Advanced Virgo (AdVirgo)'s third observing run (O3).
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:01:09 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Tsukamoto",
"Seiya",
""
],
[
"Toivonen",
"Andrew",
""
],
[
"Griffin",
"Holton",
""
],
[
"Raghuvanshi",
"Avyukt",
""
],
[
"Averill",
"Megan",
""
],
[
"Kerkow",
"Frank",
""
],
[
"Coughlin",
"Michael W.",
""
],
[
"Chan",
"Man Leong",
""
],
[
"Singer",
"Leo",
""
]
] |
2508.14243
|
Thomas Mulkey
|
T. Mulkey, K. N. Lopez, C. D. Hughes, B. Hill, M. Van Meter, H. Wijeratne, J. C. Long, M. Sarsour, W. M. Snow, K. Li, R. Parajuli, S. Samiei, D. V. Baxter, M. Luxnat, Y. Zhang, C. Jiang, E. Stringfellow, J. Torres, and R. Hobbs
|
New Upper Bounds on Exotic Neutron Spin-Electron Spin Interactions via Neutron Spin Rotation Measurements in a Compensated Ferrimagnet
| null | null | null | null |
nucl-ex cond-mat.mtrl-sci hep-ex quant-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We report a search for exotic spin-spin interactions between neutrons and electrons which could signal new physics beyond the Standard Model using slow neutron polarimetric imaging through a dense medium of polarized electrons. Our dense polarized electron target is a ferrimagnet held at its magnetic compensation temperature, which realizes a polarized electron ensemble with zero net magnetization. We sought the spin rotation of transversely polarized neutrons from a neutron spin-electron spin interaction of the form $V_2=-g_A^eg_A^n\frac{\hbar c}{4\pi}\vec\sigma_e\cdot\vec\sigma_n\frac{e^{-r/\lambda_c}}{r}$, where $g_{A}^{e}$ and $g_{A}^{n}$ are the electron and neutron axial couplings, $\vec{\sigma_e}$ and $\vec{\sigma_n}$ are the electron and neutron spin, and $\lambda_c$ is the interaction range for an exotic axial vector interaction from massive spin-1 boson exchange of mass $\hbar c/\lambda_c$. The resulting average neutron spin rotation angle per unit length, $\frac{d\bar{\phi}_{F5}}{dz}=[0.41\pm6.30\ (stat.)\pm4.4\ (sys.)]\times10^{-3}$ rad/m, is consistent with zero. Our novel approach improves the previous upper limits on the coupling constant product $g_A^eg_A^n$ by several orders of magnitude in the poorly explored $10^{-8}\leq\lambda_c\leq10^{-2}$ range.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:01:14 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Mulkey",
"T.",
""
],
[
"Lopez",
"K. N.",
""
],
[
"Hughes",
"C. D.",
""
],
[
"Hill",
"B.",
""
],
[
"Van Meter",
"M.",
""
],
[
"Wijeratne",
"H.",
""
],
[
"Long",
"J. C.",
""
],
[
"Sarsour",
"M.",
""
],
[
"Snow",
"W. M.",
""
],
[
"Li",
"K.",
""
],
[
"Parajuli",
"R.",
""
],
[
"Samiei",
"S.",
""
],
[
"Baxter",
"D. V.",
""
],
[
"Luxnat",
"M.",
""
],
[
"Zhang",
"Y.",
""
],
[
"Jiang",
"C.",
""
],
[
"Stringfellow",
"E.",
""
],
[
"Torres",
"J.",
""
],
[
"Hobbs",
"R.",
""
]
] |
2508.14244
|
Marissa Loving
|
Marissa Loving
|
Volumes of end-periodic mapping tori
|
31 pages, 33 figures; This expository paper builds on lecture notes from a minicourse entitled, "Infinite-type surfaces, end-periodic homeomorphisms, and the geometry of 3-manifolds", which was given by the author as part of the 2023 Riverside Workshop on Geometric Group Theory
| null | null | null |
math.GT math.DS
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
In this expository paper, we provide an intuition and illustration-driven overview of two recent results that tie the dynamics of certain homeomorphisms of infinite-type surfaces, called end-periodic homeomorphisms, to the geometry of their associated (compactified) mapping tori. These results are analogues of a theorem of Brock in the finite-type setting for mapping tori of pseudo-Anosov homeomorphisms.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:02:41 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Loving",
"Marissa",
""
]
] |
2508.14245
|
Shuting Du
|
Shuting Du, Mohamed Ibrahim, Zishen Wan, Luqi Zheng, Boheng Zhao, Zhenkun Fan, Che-Kai Liu, Tushar Krishna, Arijit Raychowdhury, Haitong Li
|
Cross-Layer Design of Vector-Symbolic Computing: Bridging Cognition and Brain-Inspired Hardware Acceleration
| null | null | null | null |
cs.AR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Vector Symbolic Architectures (VSAs) have been widely deployed in various cognitive applications due to their simple and efficient operations. The widespread adoption of VSAs has, in turn, spurred the development of numerous hardware solutions aimed at optimizing their performance. Despite these advancements, a comprehensive and unified discourse on the convergence of hardware and algorithms in the context of VSAs remains somewhat limited. The paper aims to bridge the gap between theoretical software-level explorations and the development of efficient hardware architectures and emerging technology fabrics for VSAs, providing insights from the co-design aspect for researchers from either side. First, we introduce the principles of vector-symbolic computing, including its core mathematical operations and learning paradigms. Second, we provide an in-depth discussion on hardware technologies for VSAs, analyzing analog, mixed-signal, and digital circuit design styles. We compare hardware implementations of VSAs by carrying out detailed analysis of their performance characteristics and tradeoffs, allowing us to extract design guidelines for the development of arbitrary VSA formulations. Third, we discuss a methodology for cross-layer design of VSAs that identifies synergies across layers and explores key ingredients for hardware/software co-design of VSAs. Finally, as a concrete demonstration of this methodology, we propose the first in-memory computing hierarchical cognition hardware system, showcasing the efficiency, flexibility, and scalability of this co-design approach. The paper concludes with a discussion of open research challenges for future explorations.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:02:50 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Du",
"Shuting",
""
],
[
"Ibrahim",
"Mohamed",
""
],
[
"Wan",
"Zishen",
""
],
[
"Zheng",
"Luqi",
""
],
[
"Zhao",
"Boheng",
""
],
[
"Fan",
"Zhenkun",
""
],
[
"Liu",
"Che-Kai",
""
],
[
"Krishna",
"Tushar",
""
],
[
"Raychowdhury",
"Arijit",
""
],
[
"Li",
"Haitong",
""
]
] |
2508.14246
|
Youqi Song
|
Youqi Song (STAR Collaboration)
|
Probing hadronization with the charge correlator ratio in $p$+$p$, $Ru$+$Ru$ and $Zr$+$Zr$ collisions at STAR
| null | null | null | null |
nucl-ex
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
The parton-to-hadron transition, known as hadronization, is dominated by non-perturbative Quantum Chromodynamics (QCD) effects and thus challenging to study from first-principle calculations. On the other hand, experimental studies of observables sensitive to hadronization could provide valuable input. The charge correlator ratio $r_c$ studies the charge correlation between the leading two hadrons in jets and is sensitive to hadronization effects. With data taken at $\sqrt{s_{\rm{NN}}}=200$ GeV at STAR, we measure $r_c$ in $p$+$p$ collisions to probe for string-like fragmentation, and in $Ru$+$Ru$ and $Zr$+$Zr$ collisions to probe for potential modification to hadronization in the Quark-Gluon Plasma (QGP). These measurements, compared with various model predictions, are expected to distinguish between phenomenological model descriptions for hadronization in vacuum, and provide insight into jet-medium interaction in the QGP.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:07:24 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Song",
"Youqi",
"",
"STAR Collaboration"
]
] |
2508.14247
|
Saswata Jana
|
Saswata Jana, Subhajit Pramanick, Adri Bhattacharya and Partha Sarathi Mandal
|
Time-optimal Asynchronous Minimal Vertex Covering by Myopic Robots
| null | null | null | null |
cs.DC
|
http://creativecommons.org/licenses/by/4.0/
|
In a connected graph with an autonomous robot swarm with limited visibility, it is natural to ask whether the robots can be deployed to certain vertices satisfying a given property using only local knowledge. This paper affirmatively answers the question with a set of \emph{myopic} (finite visibility range) luminous robots with the aim of \emph{filling a minimal vertex cover} (MVC) of a given graph $G = (V, E)$. The graph has special vertices, called \emph{doors}, through which robots enter sequentially. Starting from the doors, the goal of the robots is to settle on a set of vertices that forms a minimal vertex cover of $G$ under the asynchronous ($\mathcal{ASYNC}$) scheduler. We are also interested in achieving the \emph{minimum vertex cover} (MinVC, which is NP-hard \cite{Karp1972} for general graphs) for a specific graph class using the myopic robots. We establish lower bounds on the visibility range for the robots and on the time complexity (which is $\Omega(|E|)$). We present two algorithms for trees: one for single door, which is both time and memory-optimal, and the other for multiple doors, which is memory-optimal and achieves time-optimality when the number of doors is a constant. Interestingly, our technique achieves MinVC on trees with a single door. We then move to the general graph, where we present two algorithms, one for the single door and the other for the multiple doors with an extra memory of $O(\log \Delta)$ for the robots, where $\Delta$ is the maximum degree of $G$. All our algorithms run in $O(|E|)$ epochs.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:07:51 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Jana",
"Saswata",
""
],
[
"Pramanick",
"Subhajit",
""
],
[
"Bhattacharya",
"Adri",
""
],
[
"Mandal",
"Partha Sarathi",
""
]
] |
2508.14248
|
Marco Polver
|
Marco Polver, Daniel Limon, Fabio Previdi, Antonio Ferramosca
|
Robust tracking MPC for perturbed nonlinear systems -- Extended version
| null | null | null | null |
eess.SY cs.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents a novel robust predictive controller for constrained nonlinear systems that is able to track piece-wise constant setpoint signals. The tracking model predictive controller presented in this paper extends the nonlinear MPC for tracking to the more complex case of nonlinear systems subject to bounded and not necessarily additive perturbations. The optimal control problem that is solved at each step penalizes the deviation of the predicted nominal system trajectory from an artificial reference, which is added as a decision variable, as well as the distance between the artificial reference and the setpoint. Robust feasibility is ensured by imposing conservative constraints that take into account the effect of uncertainties and convergence to a neighborhood of any feasible setpoint is guaranteed by means of an appropriate terminal cost and an extended stabilizing terminal constraint. In the case of unreachable setpoints, convergence to a neighborhood of the optimal reachable steady output is also proved.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:08:35 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Polver",
"Marco",
""
],
[
"Limon",
"Daniel",
""
],
[
"Previdi",
"Fabio",
""
],
[
"Ferramosca",
"Antonio",
""
]
] |
2508.14249
|
Matthias Hetzenberger
|
Matthias Hetzenberger, Georg Moser, Florian Zuleger
|
To Zip Through the Cost Analysis of Probabilistic Programs
| null | null | null | null |
cs.LO
|
http://creativecommons.org/licenses/by/4.0/
|
Probabilistic programming and the formal analysis of probabilistic algorithms are active areas of research, driven by the widespread use of randomness to improve performance. While functional correctness has seen substantial progress, automated reasoning about expected runtime remains comparatively limited. In this work, we address this challenge by introducing a refinement-typed probability monad in Liquid Haskell. Our monad enables automated reasoning about expected values and costs by encoding probabilistic behaviour directly in types. Initially defined for discrete distributions over finite support, it is extended to support infinite distributions via an axiomatic approach. By leveraging Liquid Haskell's SMT-based refinement type checking, our framework provides a high degree of automation. We evaluate our approach through four case studies: meldable heaps, coupon collector, randomised quicksort, and zip trees. The first two demonstrate automation with minimal annotation overhead. The latter two showcase how our monad integrates with interactive proofs, including the first formal verification of the expected runtime of zip trees.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:20:15 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Hetzenberger",
"Matthias",
""
],
[
"Moser",
"Georg",
""
],
[
"Zuleger",
"Florian",
""
]
] |
2508.14250
|
Vsevolod Ivanov
|
Vsevolod R. Ivanov
|
Inflationary Slow-Roll Parameters in the Jordan Frame for Cosmological F(R) Gravity Models
|
10 pages, 2 figures
| null | null | null |
gr-qc astro-ph.CO
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
$F(R)$ inflationary models are analyzed without the Weyl transformation to the Einstein frame. Sufficient conditions for existence of global inflationary attractors in $F(R)$ models are provided. Following that, the procedure for calculating inflationary observables completely in the original, Jordan, frame is given. The described procedure is used to analyze a new $F(R)$ inflationary model (a one-parameter deformation of the Starobinsky model), which is shown to provide predictions in good agreement with the recent ACT data.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:20:23 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Ivanov",
"Vsevolod R.",
""
]
] |
2508.14251
|
Olivier Sarbach
|
Carlos Gabarrete and Ana Laura Garc\'ia-Perciante and Olivier Sarbach
|
The Chapman-Enskog approximation for a relativistic charged gas in the trace-fixed particle frame
|
29 pages, no figures
| null | null | null |
gr-qc cond-mat.stat-mech hep-th
|
http://creativecommons.org/licenses/by/4.0/
|
In this article we provide a thorough discussion of the first-order Chapman-Enskog approximation for a relativistic simple charged gas propagating in an arbitrary fixed background spacetime and electromagnetic field. Motivated by a rigorous study of the linearized collision operator, we argue that the most natural thermodynamic frame to describe dissipative relativistic fluids is the trace-fixed particle frame. This frame determines the state variables by requiring compatibility of the first few moments of the one-particle distribution function with those of the J\"uttner distribution. The resulting constitutive relations are derived, and it is shown that the corresponding transport coefficients are, in fact, frame-independent if suitably defined. Moreover, we discuss the representation freedom for these relations within the microscopic approach. As shown in previous work, when this freedom is suitably exploited, the resulting fluid theory gives rise to strongly hyperbolic and causal evolution equations for which global equilibria are stable.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:21:18 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Gabarrete",
"Carlos",
""
],
[
"García-Perciante",
"Ana Laura",
""
],
[
"Sarbach",
"Olivier",
""
]
] |
2508.14252
|
Jeffrey Brock
|
Jeffrey A. Brock, Aleksandr Kurenkov, Ale\v{s} Hrabec, and Laura J. Heyderman
|
Engineering and exploiting self-driven domain wall motion in ferrimagnets for neuromorphic computing applications
|
22 pages, 4 figures, 5 supporting figures
| null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Magnetic domain wall motion has recently garnered significant interest as a physical mechanism to enable energy-efficient, next-generation brain-inspired computing architectures. However, realizing all behaviors required for neuromorphic computing within standard material systems remains a significant challenge, as these functionalities often rely on competing interactions. Here, we demonstrate how spontaneous domain wall motion in response to locally engineered lateral exchange coupling in transition metal-rare earth ferrimagnets can be leveraged to achieve numerous neuromorphic computing functionalities in devices with minimal complexity. Through experiments and micromagnetic simulations, we show how tuning the feature size, material composition, and chiral interaction strength controls the speed of self-driven domain wall motion. When integrated with spin-orbit torque, this control gives rise to behaviors essential for neuromorphic computing, including leaky integration and passive resetting of artificial neuron potential. These results establish locally engineered ferrimagnets as a tunable, scalable, and straightforward platform for domain wall-based computing architectures.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:22:12 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Brock",
"Jeffrey A.",
""
],
[
"Kurenkov",
"Aleksandr",
""
],
[
"Hrabec",
"Aleš",
""
],
[
"Heyderman",
"Laura J.",
""
]
] |
2508.14253
|
Kelly Holley-Bockelmann
|
Kelly Holley-Bockelmann (1,2), Fazeel Khan (3,4), Isaiah Williams (1), Jaelyn Roth (1), Michael Rizzo Smith (1), Kaitlin Porter (1), Jillian Bellovary (5,6,7), Andrea Derdzinski (2), Andrea Macci\`o (3,4) ((1) Department of Physics and Astronomy, Vanderbilt University, Nashville, TN (2) Department of Physics, Fisk University, Nashville, TN (3) New York University Abu Dhabi, United Arab Emirates (4) Center for Astrophysics and Space Science (CASS), New York University Abu Dhabi, UAE (5) Department of Physics, Queensborough Community College, Bayside, NY (6) Department of Astrophysics, American Museum of Natural History, New York, NY (7) Astrophysics Program, CUNY Graduate Center, New York, NY)
|
Handy Relation Between Binary Black Hole Merger Times and Host Galaxy Properties
|
12 pages, submitted to ApJ -- comments welcome, especially for neglected references!
| null | null | null |
astro-ph.GA
|
http://creativecommons.org/licenses/by/4.0/
|
Over the past 15 years, the evidence has clearly demonstrated that massive black hole (MBH) binary merger timescales depend strongly on the structural and kinematic properties of their host galaxy. Stellar density, gas content, shape and kinematics all play a role, combining in non-linear ways to effect the evolution of the binary. The binary properties themselves, such as eccentricity, mass ratio, and orbital plane, all matter as well. This makes it nontrivial to estimate accurate cosmological MBH binary merger rates, or to generate merger rate ranges that reflect the distribution of galaxy hosts and orbits. Using an extensive set of high-resolution direct N-body simulations in which the shape, structure, and kinematics of each galaxy host are directly informed by observations, we map out MBH binary merger timescales over a range of galaxy hosts and MBH binary orbits. This yields a convenient set of scaling relations to determine MBH binary merger timescales -- and the range of merger timescales -- as functions of basic observables. Such scaling relations can be readily employed as a subgrid model in cosmological or semi-analytic studies, for example, to model event rates for LISA or pulsar timing.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:22:33 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Holley-Bockelmann",
"Kelly",
"",
"Department of Physics and Astronomy, Vanderbilt University, Nashville, TN",
"Department of Physics, Fisk University, Nashville, TN"
],
[
"Khan",
"Fazeel",
"",
"New York University Abu Dhabi, United Arab Emirates",
"Center for Astrophysics and Space Science"
],
[
"Williams",
"Isaiah",
"",
"Department of Physics and Astronomy, Vanderbilt University, Nashville, TN"
],
[
"Roth",
"Jaelyn",
"",
"Department of Physics and Astronomy, Vanderbilt University, Nashville, TN"
],
[
"Smith",
"Michael Rizzo",
"",
"Department of Physics and Astronomy, Vanderbilt University, Nashville, TN"
],
[
"Porter",
"Kaitlin",
"",
"Department of Physics and Astronomy, Vanderbilt University, Nashville, TN"
],
[
"Bellovary",
"Jillian",
"",
"Department of Physics, Queensborough Community College, Bayside, NY",
"Department of Astrophysics, American Museum of Natural History, New York, NY",
"Astrophysics Program, CUNY Graduate Center, New York, NY"
],
[
"Derdzinski",
"Andrea",
"",
"Department of Physics, Fisk University, Nashville, TN"
],
[
"Macciò",
"Andrea",
"",
"New York University Abu Dhabi, United Arab Emirates",
"Center for Astrophysics and Space Science"
]
] |
2508.14254
|
Amey Khanolkar
|
Amey Khanolkar, Saqeeb Adnan, Md Minaruzzaman, Linu Malakkal, Darren B. Thomson, David B. Turner, J. Matthew Mann, David H. Hurley, and Marat Khafizov
|
Lattice anharmonicity effects in fluorite oxide single crystals and anomalous increase in phonon lifetime in ceria at elevated temperature
| null | null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
We investigate the temperature dependence of the frequency and linewidth of the triply-degenerate T$_{2g}$ zone-centered optical phonon in flux-grown ceria and hydrothermally-synthesized thoria single crystals from room temperature to 1273 K using Raman spectroscopy. Both crystals exhibit an expected increase in the phonon linewidth with temperature due to enhanced phonon-phonon scattering. However, ceria displays an anomalous linewidth reduction in the temperature range of 1023-1123 K. First-principles phonon linewidth calculations considering cubic and quartic phonon interactions within temperature-independent phonon dispersion fail to describe this anomaly. A parameterization of the temperature-dependent second order interatomic force constants based on previously reported phonon dispersion measured at room and high temperatures, predicts a deviation from the monotonic linewidth increase, albeit at temperatures lower than those observed experimentally for ceria. The qualitative agreement in the trend of temperature-dependent linewidth suggests that lattice anharmonicity-induced phonon renormalization plays a role in phonon lifetime. Specifically, a change in the overlap between softened acoustic and optical branches in the dispersion curve reduces the available phonon scattering phase space of the Raman active mode at the zone center, leading to an increased phonon lifetime within a narrow temperature interval. These findings provide new insights into higher-order anharmonic interactions in ceria and thoria, motivating further investigations into the role of anharmonicity-induced phonon renormalization on phonon lifetimes at high temperatures.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:22:59 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Khanolkar",
"Amey",
""
],
[
"Adnan",
"Saqeeb",
""
],
[
"Minaruzzaman",
"Md",
""
],
[
"Malakkal",
"Linu",
""
],
[
"Thomson",
"Darren B.",
""
],
[
"Turner",
"David B.",
""
],
[
"Mann",
"J. Matthew",
""
],
[
"Hurley",
"David H.",
""
],
[
"Khafizov",
"Marat",
""
]
] |
2508.14255
|
Haotian Xu
|
Haotian Xu, Tsui-Wei Weng, Lam M. Nguyen, Tengfei Ma
|
Graph Concept Bottleneck Models
| null | null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Concept Bottleneck Models (CBMs) provide explicit interpretations for deep neural networks through concepts and allow intervention with concepts to adjust final predictions. Existing CBMs assume concepts are conditionally independent given labels and isolated from each other, ignoring the hidden relationships among concepts. However, the set of concepts in CBMs often has an intrinsic structure where concepts are generally correlated: changing one concept will inherently impact its related concepts. To mitigate this limitation, we propose GraphCBMs: a new variant of CBM that facilitates concept relationships by constructing latent concept graphs, which can be combined with CBMs to enhance model performance while retaining their interpretability. Our experiment results on real-world image classification tasks demonstrate Graph CBMs offer the following benefits: (1) superior in image classification tasks while providing more concept structure information for interpretability; (2) able to utilize latent concept graphs for more effective interventions; and (3) robust in performance across different training and architecture settings.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:23:18 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Xu",
"Haotian",
""
],
[
"Weng",
"Tsui-Wei",
""
],
[
"Nguyen",
"Lam M.",
""
],
[
"Ma",
"Tengfei",
""
]
] |
2508.14256
|
Joel Hurtado
|
Joel Hurtado, Benjamin Gerard, Cesar Laguna, Dominic Francisco Sanchez
|
Jupyter simulations Using NGS Intensity with PyWFS for REDWOODS
|
submitted to Proceedings of SPIE 2025 Optics + Photonics, Astronomical Applications, Paper number 13627-75
| null | null | null |
astro-ph.IM
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
REDWOODS on ShaneAO at Lick Observatory implements a second-stage, 3-sided reflective pyramid wavefront sensor (PWFS), which, under low-light conditions, offers an improved signal-to-noise ratio for deformable mirror commands to correct dynamic atmospheric aberrations. We modeled the REDWOODS limiting natural guide star magnitude. We also investigated performance tradeoffs between two PWFS modes that are available to REDWOODS, one with higher bandwidth error and another with higher aliasing error. We also implemented an experimental setup to image one of the REDWOODS PWFS masks.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:25:17 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Hurtado",
"Joel",
""
],
[
"Gerard",
"Benjamin",
""
],
[
"Laguna",
"Cesar",
""
],
[
"Sanchez",
"Dominic Francisco",
""
]
] |
2508.14257
|
Jonathan Zong
|
Jonathan Zong
|
Using Real Names of Disabled Participant-Contributors to Practice Citational Justice in Accessibility
|
4 pages, 0 figures. IEEE VIS Accessibility Workshop
| null | null | null |
cs.HC
|
http://creativecommons.org/licenses/by/4.0/
|
In accessibility research involving human subjects, researchers conventionally anonymize their research participants to protect privacy. However, a lack of intentionality about who to publicly acknowledge for intellectual contributions to research can lead to the erasure of disabled individuals' work and knowledge. In this paper, I propose identifying disabled research participants by name (with consent) as a practice of citational justice. I share observations from examples of this practice in accessible visualization research, and offer considerations for when it may be appropriate to de-anonymize. Intentional practices of citation offer researchers an opportunity to acknowledge the expertise and intellectual contributions of disabled people in our communities.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:26:54 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zong",
"Jonathan",
""
]
] |
2508.14258
|
Daegyun Choi
|
Daegyun Choi, Alhim Vera, Donghoon Kim
|
Adapting Biological Reflexes for Dynamic Reorientation in Space Manipulator Systems
|
18 pages, 11 figures, 2025 AAS/AIAA Astrodynamics Specialist Conference
| null | null | null |
cs.RO physics.bio-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Robotic arms mounted on spacecraft, known as space manipulator systems (SMSs), are critical for enabling on-orbit assembly, satellite servicing, and debris removal. However, controlling these systems in microgravity remains a significant challenge due to the dynamic coupling between the manipulator and the spacecraft base. This study explores the potential of using biological inspiration to address this issue, focusing on animals, particularly lizards, that exhibit mid-air righting reflexes. Based on similarities between SMSs and these animals in terms of behavior, morphology, and environment, their air-righting motion trajectories are extracted from high-speed video recordings using computer vision techniques. These trajectories are analyzed within a multi-objective optimization framework to identify the key behavioral goals and assess their relative importance. The resulting motion profiles are then applied as reference trajectories for SMS control, with baseline controllers used to track them. The findings provide a step toward translating evolved animal behaviors into interpretable, adaptive control strategies for space robotics, with implications for improving maneuverability and robustness in future missions.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:26:55 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Choi",
"Daegyun",
""
],
[
"Vera",
"Alhim",
""
],
[
"Kim",
"Donghoon",
""
]
] |
2508.14259
|
Jacob Deutsch
|
Jacob Deutsch
|
Finding Inter-species Associations on Large Citizen Science Datasets
| null | null | null | null |
q-bio.PE
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Determining associations among different species from citizen science databases is challenging due to observer behavior and intrinsic density variations that give rise to correlations that do not imply species associations. This paper introduces a method that can efficiently analyze large datasets to extract likely species associations. It tiles space into small blocks chosen to be of the accuracy of the data coordinates, and reduces observations to presence/absence per tile, in order to compute pairwise overlaps. It compares these overlaps with a spatial Poisson process that serves as a null model. For each species $i$, an expected overlap $\mu_i$ is estimated by averaging normalized overlaps over other species in the same vicinity. This gives a $z$-score for significance of a species-species association and a correlation index for the strength of this association. This was tested on $874,263$ iNaturalist observations spanning $15,975$ non-avian taxa in the Santa Cruz, California region ($\approx 4.68\times10^{6}$ tiles). The method recovers well-known insect host-plant obligate relationships, particularly many host-gall relationships, as well as the relationship between Yerba Santa Beetles and California Yerba Santa. This approach efficiently finds associations on $\sim10^{8}$ species pairs on modest hardware, filtering correlations arising from heterogeneous spatial prevalence and user artifacts. It produces a ranked shortlist of ecological interactions that can be further pursued. Extensions to this method are possible, such as investigating the effects of time and elevation. It could also be useful in the determination of microhabitats and biomes.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:31:08 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Deutsch",
"Jacob",
""
]
] |
2508.14260
|
Marcelo Alberto Cort\'es Vergara
|
M.C. Vergara, A. Askar, F. Flammini Dotti, D.R.G. Schleicher, A. Escala, R. Spurzem, M. Giersz, J. Hurley, M. Arca Sedda, and N. Neumayer
|
Efficient black hole seed formation in low metallicity and dense stellar clusters with implications for JWST sources
|
16 pages, 12 figures
| null | null | null |
astro-ph.GA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Recent observations with the James Webb Space Telescope (JWST) reveal young massive clusters (YMCs) as key building blocks of early galaxies. They are not only important constituents of galaxies, but also potential birthplaces of very massive stars (VMSs) and black hole (BH) seeds. We explore stellar dynamics in extremely dense clusters with initial half-mass densities of $\rho_h \gtrsim 10^8M_\odot{\rm pc}^{-3}$ at very low metallicity, comparable to some of the densest clusters seen by JWST. Using direct N-body and Monte Carlo simulations with stellar evolution, we show that VMS formation through collisions is unavoidable, with final masses reaching $5\times10^3$ to $4\times10^4M_\odot$. These results support the existence of a critical mass scale above which collisions become highly efficient, enabling the formation of VMSs and intermediate-mass BHs (IMBHs). Our models, using nbody6++gpu and MOCCA with updated SSE/BSE routines, show that dense clusters rapidly form VMSs via stellar bombardment. The VMSs then collapse into BH seeds of a few $10^3$ to $10^4M_\odot$ in less than 4 Myr. We identify a critical mass-density threshold beyond which clusters undergo runaway collisions that yield massive BH seeds. For typical YMCs detected by JWST, efficiencies up to 10% are expected, implying BH masses up to $10^5M_\odot$ if formed via collisions. We predict a scaling relation for BH mass, $\log(M_{\rm BH}/M_\odot)=-0.76+0.76\log(M/M_\odot)$. Frequent VMS formation may also explain the high nitrogen abundance observed in galaxies at high redshift.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:32:30 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Vergara",
"M. C.",
""
],
[
"Askar",
"A.",
""
],
[
"Dotti",
"F. Flammini",
""
],
[
"Schleicher",
"D. R. G.",
""
],
[
"Escala",
"A.",
""
],
[
"Spurzem",
"R.",
""
],
[
"Giersz",
"M.",
""
],
[
"Hurley",
"J.",
""
],
[
"Sedda",
"M. Arca",
""
],
[
"Neumayer",
"N.",
""
]
] |
2508.14261
|
Meet Udeshi
|
Meet Udeshi and Venkata Sai Charan Putrevu and Prashanth Krishnamurthy and Ramesh Karri and Farshad Khorrami
|
SaMOSA: Sandbox for Malware Orchestration and Side-Channel Analysis
| null | null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by/4.0/
|
Cyber-attacks on operational technology (OT) and cyber-physical systems (CPS) have increased tremendously in recent years with the proliferation of malware targeting Linux-based embedded devices of OT and CPS systems. Comprehensive malware detection requires dynamic analysis of execution behavior in addition to static analysis of binaries. Safe execution of malware in a manner that captures relevant behaviors via side-channels requires a sandbox environment. Existing Linux sandboxes are built for specific tasks, only capture one or two side-channels, and do not offer customization for different analysis tasks. We present the SaMOSA Linux sandbox that allows emulation of Linux malwares while capturing time-synchronized side-channels from four sources. SaMOSA additionally provides emulation of network services via FakeNet, and allows orchestration and customization of the sandbox environment via pipeline hooks. In comparison to existing Linux sandboxes, SaMOSA captures more side-channels namely system calls, network activity, disk activity, and hardware performance counters. It supports three architectures predominantly used in OT and CPS namely x86-64, ARM64, and PowerPC 64. SaMOSA fills a gap in Linux malware analysis by providing a modular and customizable sandbox framework that can be adapted for many malware analysis tasks. We present three case studies of three different malware families to demonstrate the advantages of SaMOSA.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:41:24 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Udeshi",
"Meet",
""
],
[
"Putrevu",
"Venkata Sai Charan",
""
],
[
"Krishnamurthy",
"Prashanth",
""
],
[
"Karri",
"Ramesh",
""
],
[
"Khorrami",
"Farshad",
""
]
] |
2508.14262
|
Graeme Pleasance
|
Graeme Pleasance, Erik Aurell, Francesco Petruccione
|
A high-temperature limit penalizing high-frequency quantum fluctuations
|
12 pages, 4 figures
| null | null | null |
quant-ph
|
http://creativecommons.org/licenses/by/4.0/
|
We revisit the Caldeira-Leggett model of quantum Brownian motion with Ohmic spectral density, and derive an additional contribution to the decoherence kernel in a new high-temperature limit at arbitrarily large cut-off frequency. This contribution reveals a novel mechanism for the classicalization of high-frequency quantum fluctuations. We further demonstrate that it leads to a Markovian master equation that is in guaranteed Lindblad form. Our approach considers in detail the behavior of the decoherence kernel at both the initial and final times of the process on the time scale of the bath memory.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:41:32 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Pleasance",
"Graeme",
""
],
[
"Aurell",
"Erik",
""
],
[
"Petruccione",
"Francesco",
""
]
] |
2508.14263
|
Michael Borinsky
|
Michael Borinsky
|
Tropicalized quantum field theory and global tropical sampling
|
36 pages
| null | null | null |
math-ph hep-th math.AG math.CO math.MP
|
http://creativecommons.org/licenses/by/4.0/
|
We explain how to tropicalize scalar quantum field theory and show that tropicalized massive scalar quantum field theory is exactly solvable. This exact solution manifests as a non-linear recursion equation fulfilled by the expansion coefficients of the quantum effective action. Geometrically, this recursion computes specific volumes of moduli spaces of metric graphs and is thereby analogous to Mirzakhani's volume recursions on the moduli space of curves. Building on this exact solution, we construct an algorithm that samples points from the moduli space of graphs approximately proportional to their perturbative contribution. Remarkably, this algorithm requires only polynomial time and memory, suggesting that perturbative quantum field theory computations lie in the polynomial-time complexity class, while all known algorithms for evaluating individual Feynman integrals are exponential in time and memory. To demonstrate the capabilities of the algorithm, we evaluate the primitive contribution to the $\phi^4$ beta function at 50 loops with a proof-of-concept implementation.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:52:47 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Borinsky",
"Michael",
""
]
] |
2508.14264
|
Thanh-Dat Truong
|
Thanh-Dat Truong, Huu-Thien Tran, Tran Thai Son, Bhiksha Raj, Khoa Luu
|
Directed-Tokens: A Robust Multi-Modality Alignment Approach to Large Language-Vision Models
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Large multimodal models (LMMs) have gained impressive performance due to their outstanding capability in various understanding tasks. However, these models still suffer from some fundamental limitations related to robustness and generalization due to the alignment and correlation between visual and textual features. In this paper, we introduce a simple but efficient learning mechanism for improving the robust alignment between visual and textual modalities by solving shuffling problems. In particular, the proposed approach can improve reasoning capability, visual understanding, and cross-modality alignment by introducing two new tasks: reconstructing the image order and the text order into the LMM's pre-training and fine-tuning phases. In addition, we propose a new directed-token approach to capture visual and textual knowledge, enabling the capability to reconstruct the correct order of visual inputs. Then, we introduce a new Image-to-Response Guided loss to further improve the visual understanding of the LMM in its responses. The proposed approach consistently achieves state-of-the-art (SoTA) performance compared with prior LMMs on academic task-oriented and instruction-following LMM benchmarks.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:53:24 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Truong",
"Thanh-Dat",
""
],
[
"Tran",
"Huu-Thien",
""
],
[
"Son",
"Tran Thai",
""
],
[
"Raj",
"Bhiksha",
""
],
[
"Luu",
"Khoa",
""
]
] |
2508.14265
|
Sadmir Kudin
|
Sadmir Kudin, Enes Pasalic, Alexandr Polujan, Fengrong Zhang and Haixia Zhao
|
Almost Maiorana-McFarland bent functions
| null | null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
In this article, we study bent functions on $\mathbb{F}_2^{2m}$ of the form $f(x,y) = x \cdot \phi(y) + h(y)$, where $x \in \mathbb{F}_2^{m-1} $ and $ y \in \mathbb{F}_2^{m+1}$, which form the generalized Maiorana-McFarland class (denoted by ${GMM}_{m+1}$) and are referred to as almost Maiorana-McFarland bent functions. We provide a complete characterization of the bent property for such functions and determine their duals. Specifically, we show that $f$ is bent if and only if the mapping $\phi $ partitions $ \mathbb{F}_2^{m+1}$ into 2-dimensional affine subspaces, on each of which the function $ h $ has odd weight. We investigate which properties of mappings $\phi \colon \mathbb{F}_2^{m+1} \to \mathbb{F}_2^{m-1}$ lead to bent functions of the form $ f(x,y) = x \cdot \phi(y) + h(y) $ both inside and outside ${M}^\# $ and provide construction methods for suitable Boolean functions $ h $ on $\mathbb{F}_2^{m+1}$. We present a simple algorithm for constructing partitions of the vector space $\mathbb{F}_2^{m+1}$ together with appropriate Boolean functions $ h $ that generate bent functions outside ${M}^\# $. When $ 2m = 8 $, we explicitly identify many such partitions that produce at least $ 2^{78} $ distinct bent functions on $\mathbb{F}_2^8$ that do not belong to ${M}^\# $, thereby generating more bent functions outside ${M}^\#$ than the total number of 8-variable bent functions in ${M}^\#$. Additionally, we demonstrate that concatenating four almost Maiorana-McFarland bent functions outside ${M}^\# $, can result in a bent function ${M}^\# $. This finding answers an open problem posed recently in Kudin et al. (IEEE Trans. Inf. Theory 71(5): 3999-4011, 2025). Conversely, using a similar approach to concatenate four functions each in ${M}^\#$, we generate bent functions that are provably outside ${M}^\#$.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:53:27 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Kudin",
"Sadmir",
""
],
[
"Pasalic",
"Enes",
""
],
[
"Polujan",
"Alexandr",
""
],
[
"Zhang",
"Fengrong",
""
],
[
"Zhao",
"Haixia",
""
]
] |
2508.14266
|
Rizwan Ahamed
|
Rizwan Ahamed, Annahita Amireskandari, Joel Palko, Carol Laxson, Binod Bhattarai, Prashnna Gyawali
|
Effect of Data Augmentation on Conformal Prediction for Diabetic Retinopathy
|
3rd Workshop in Data Engineering in Medical Imaging (DEMI), MICCAI-2025 Workshop
| null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
The clinical deployment of deep learning models for high-stakes tasks such as diabetic retinopathy (DR) grading requires demonstrable reliability. While models achieve high accuracy, their clinical utility is limited by a lack of robust uncertainty quantification. Conformal prediction (CP) offers a distribution-free framework to generate prediction sets with statistical guarantees of coverage. However, the interaction between standard training practices like data augmentation and the validity of these guarantees is not well understood. In this study, we systematically investigate how different data augmentation strategies affect the performance of conformal predictors for DR grading. Using the DDR dataset, we evaluate two backbone architectures -- ResNet-50 and a Co-Scale Conv-Attentional Transformer (CoaT) -- trained under five augmentation regimes: no augmentation, standard geometric transforms, CLAHE, Mixup, and CutMix. We analyze the downstream effects on conformal metrics, including empirical coverage, average prediction set size, and correct efficiency. Our results demonstrate that sample-mixing strategies like Mixup and CutMix not only improve predictive accuracy but also yield more reliable and efficient uncertainty estimates. Conversely, methods like CLAHE can negatively impact model certainty. These findings highlight the need to co-design augmentation strategies with downstream uncertainty quantification in mind to build genuinely trustworthy AI systems for medical imaging.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:55:06 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Ahamed",
"Rizwan",
""
],
[
"Amireskandari",
"Annahita",
""
],
[
"Palko",
"Joel",
""
],
[
"Laxson",
"Carol",
""
],
[
"Bhattarai",
"Binod",
""
],
[
"Gyawali",
"Prashnna",
""
]
] |
2508.14267
|
Mihai-Silviu Lazorec
|
Mihai-Silviu Lazorec and Marius T\u{a}rn\u{a}uceanu
|
On the number of conjugacy classes of subgroups of a finite group
|
15 pages
| null | null | null |
math.GR
|
http://creativecommons.org/licenses/by/4.0/
|
Let $k'(G)$ and $L(G)$ be the number of conjugacy classes of subgroups and the subgroup lattice of a finite group $G$, respectively. Our objective is to study some aspects related to the ratios $d'(G)=\frac{k'(G)}{|L(G)|}$ and $d^*(G)=\min\{ d'(S) \mid S\text{ is a section of }G\}$ which measure how close is $G$ from being a Dedekind group. We prove that the set containing the values $d'(G)$, as $G$ ranges over the class of nilpotent groups, is dense in $[0, 1]$. A nilpotency criterion is obtained by proving that if $d^*(G)>\frac{2}{3}$, then $G$ is nilpotent and information on its structure is given. We also show that if $d^*(G)>\frac{4}{5}$, then $G$ is an Iwasawa group. Finally, we deduce a result which ensures that a $p$-group of order $p^n$ ($n\geq 3$) is a Dedekind group. This last result can be extended to the class of nilpotent groups and it also highlights the second maximum values of $d'$ and $d^*$ on the class of $p$-groups of order $p^n$.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:55:12 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Lazorec",
"Mihai-Silviu",
""
],
[
"Tărnăuceanu",
"Marius",
""
]
] |
2508.14268
|
Chenghui Zheng
|
Chenghui Zheng, Garvesh Raskutti
|
Comparing Model-agnostic Feature Selection Methods through Relative Efficiency
| null | null | null | null |
stat.ML cs.LG stat.ME
|
http://creativecommons.org/licenses/by/4.0/
|
Feature selection and importance estimation in a model-agnostic setting is an ongoing challenge of significant interest. Wrapper methods are commonly used because they are typically model-agnostic, even though they are computationally intensive. In this paper, we focus on feature selection methods related to the Generalized Covariance Measure (GCM) and Leave-One-Covariate-Out (LOCO) estimation, and provide a comparison based on relative efficiency. In particular, we present a theoretical comparison under three model settings: linear models, non-linear additive models, and single index models that mimic a single-layer neural network. We complement this with extensive simulations and real data examples. Our theoretical results, along with empirical findings, demonstrate that GCM-related methods generally outperform LOCO under suitable regularity conditions. Furthermore, we quantify the asymptotic relative efficiency of these approaches. Our simulations and real data analysis include widely used machine learning methods such as neural networks and gradient boosting trees.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:55:43 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zheng",
"Chenghui",
""
],
[
"Raskutti",
"Garvesh",
""
]
] |
2508.14269
|
Jinyoung Park
|
Quentin Dubroff, Jeff Kahn, Jinyoung Park
|
On the "second" Kahn--Kalai Conjecture
|
14 pages
| null | null | null |
math.CO math.PR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We make progress on a conjecture of Kahn and Kalai, the original (stronger but less general) version of what became known as the ``Kahn-Kalai Conjecture" (KKC; now a theorem of Park and Pham). This ``second" KKC concerns the threshold, $p_c(H)$, for $G_{n,p}$ to contain a copy of a given graph $H$, predicting $p_c(H) = O(p_{\mathbb E}(H)\log n)$, where $p_{\mathbb E}$ is an easy lower bound on $p_c$. What we actually show is $p_{\mathbb E}^*(H)=O(p_{\mathbb E}(H)\log ^2n)$, where $p_{\mathbb E}^*$, the fractional expectation threshold, is a larger lower bound suggested by Talagrand. When combined with Talagrand's fractional relaxation of the KKC (now a theorem of Frankston, Kahn, Narayanan and Park), this gives $p_c(H)=O(p_{\mathbb E}(H)\log^3 n)$. (The second KKC would follow similarly if one could remove the log factors from the above bound on $p_{\mathbb E}^*$.)
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:55:56 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Dubroff",
"Quentin",
""
],
[
"Kahn",
"Jeff",
""
],
[
"Park",
"Jinyoung",
""
]
] |
2508.14270
|
Ramon Cardias
|
Ramon Cardias, Hugo U. R. Strand, Anders Bergman, A. B. Klautau and Tatiana G. Rappoport
|
Real-space first-principles approach to orbitronic phenomena in metallic multilayers
|
12 pages, 5 figures
| null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/publicdomain/zero/1.0/
|
We develop a real-space first-principles method based on density functional theory to investigate orbitronic phenomena in complex materials. Using the Real-Space Linear Muffin-Tin Orbital method within the Atomic Sphere Approximation (RS-LMTO-ASA) combined with a Chebyshev polynomial expansion of the Green's functions, we compute orbital (spin) Hall transport and orbital (spin) accumulation directly in real space. The approach scales linearly with system size and naturally incorporates disorder, finite-size effects, and interface roughness. We apply the method to transition-metal-based heterostructures and demonstrate the emergence of substantial orbital (spin) accumulation, even in centrosymmetric systems. Our methodology provides a scalable and flexible framework for realistic simulations of orbital transport phenomena in complex heterostructures.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 20:58:27 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Cardias",
"Ramon",
""
],
[
"Strand",
"Hugo U. R.",
""
],
[
"Bergman",
"Anders",
""
],
[
"Klautau",
"A. B.",
""
],
[
"Rappoport",
"Tatiana G.",
""
]
] |
2508.14271
|
Saul Youssef
|
Saul Youssef
|
Pure Data Spaces
|
26 pages
| null | null | null |
cs.DC math.LO
|
http://creativecommons.org/licenses/by/4.0/
|
In a previous work, "pure data" is proposed as an axiomatic foundation for mathematics and computing, based on "finite sequence" as the foundational concept rather than based on logic or type. Within this framework, objects with mathematical meaning are "data" and collections of mathematical objects must then be associative data, called a "space." A space is then the basic collection in this framework analogous to sets in Set Theory or objects in Category Theory. A theory of spaces is developed,where spaces are studied via their semiring of endomorphisms. To illustrate these concepts, and as a way of exploring the implications of the framework, pure data spaces are "grown organically" from the substrate of pure data with minimal combinatoric definitions. Familiar objects from classical mathematics emerge this way, including natural numbers, integers, rational numbers, boolean spaces, matrix algebras, Gaussian Integers, Quaternions, and non-associative algebras like the Integer Octonions. Insights from these examples are discussed with a view towards new directions in theory and new exploration.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:01:13 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Youssef",
"Saul",
""
]
] |
2508.14272
|
Ron Folman
|
Peter Skakunenko, Daniel Folman, Yaniv Bar-Haim, Ron Folman
|
Strong Confinement of a Nanoparticle in a Needle Paul Trap: Towards Matter-Wave Interferometry with Nanodiamonds
|
arXiv admin note: text overlap with arXiv:2508.13723
| null | null | null |
quant-ph gr-qc physics.atom-ph
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Quantum mechanics (QM) and General relativity (GR), also known as the theory of gravity, are the two pillars of modern physics. A matter-wave interferometer with a massive particle, can test numerous fundamental ideas, including the spatial superposition principle - a foundational concept in QM - in completely new regimes, as well as the interface between QM and GR, e.g., testing the quantization of gravity. Consequently, there exists an intensive effort to realize such an interferometer. While several paths are being pursued, we focus on utilizing nanodiamonds as our particle, and a spin embedded in the nanodiamond together with Stern-Gerlach forces, to achieve a closed loop in space-time. There is a growing community of groups pursuing this path [1]. We are posting this technical note (as part of a series of seven such notes), to highlight our plans and solutions concerning various challenges in this ambitious endeavor, hoping this will support this growing community. In this work, we achieve strong confinement of a levitated particle, which is crucial for angular confinement, precise positioning, and perhaps also advantageous for deep cooling. We designed a needle Paul trap with a controllable distance between the electrodes, giving rise to a strong electric gradient. By combining it with an effective charging method - electrospray - we reach a trap frequency of up to 40 kHz, which is more than twice the state of the art. We believe that the designed trap could become a significant tool in the hands of the community working towards massive matter-wave interferometry. We would be happy to make more details available upon request.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:02:57 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Skakunenko",
"Peter",
""
],
[
"Folman",
"Daniel",
""
],
[
"Bar-Haim",
"Yaniv",
""
],
[
"Folman",
"Ron",
""
]
] |
2508.14273
|
Krishna Garg
|
Krishna Garg, Firoz Shaikh, Sambaran Bandyopadhyay, Cornelia Caragea
|
Let's Use ChatGPT To Write Our Paper! Benchmarking LLMs To Write the Introduction of a Research Paper
|
20 pages, 15 figures
| null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
As researchers increasingly adopt LLMs as writing assistants, generating high-quality research paper introductions remains both challenging and essential. We introduce Scientific Introduction Generation (SciIG), a task that evaluates LLMs' ability to produce coherent introductions from titles, abstracts, and related works. Curating new datasets from NAACL 2025 and ICLR 2025 papers, we assess five state-of-the-art models, including both open-source (DeepSeek-v3, Gemma-3-12B, LLaMA 4-Maverick, MistralAI Small 3.1) and closed-source GPT-4o systems, across multiple dimensions: lexical overlap, semantic similarity, content coverage, faithfulness, consistency, citation correctness, and narrative quality. Our comprehensive framework combines automated metrics with LLM-as-a-judge evaluations. Results demonstrate LLaMA-4 Maverick's superior performance on most metrics, particularly in semantic similarity and faithfulness. Moreover, three-shot prompting consistently outperforms fewer-shot approaches. These findings provide practical insights into developing effective research writing assistants and set realistic expectations for LLM-assisted academic writing. To foster reproducibility and future research, we will publicly release all code and datasets.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:11:11 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Garg",
"Krishna",
""
],
[
"Shaikh",
"Firoz",
""
],
[
"Bandyopadhyay",
"Sambaran",
""
],
[
"Caragea",
"Cornelia",
""
]
] |
2508.14274
|
Qiyi Tang
|
Mona Alluwayma, Yong Li, Sven Schewe and Qiyi Tang
|
Efficient Learning of Weak Deterministic B\"uchi Automata
|
accepted at 28th European Conference on Artificial Intelligence (ECAI 2025), 9 pages, 6 figures
| null | null | null |
cs.FL
|
http://creativecommons.org/licenses/by/4.0/
|
We present an efficient Angluin-style learning algorithm for weak deterministic B\"uchi automata (wDBAs). Different to ordinary deterministic B\"uchi and co-B\"uchi automata, wDBAs have a minimal normal form, and we show that we can learn this minimal normal form efficiently. We provide an improved result on the number of queries required and show on benchmarks that this theoretical advantage translates into significantly fewer queries: while previous approaches require a quintic number of queries, we only require quadratically many queries in the size of the canonic wDBA that recognises the target language.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:11:47 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Alluwayma",
"Mona",
""
],
[
"Li",
"Yong",
""
],
[
"Schewe",
"Sven",
""
],
[
"Tang",
"Qiyi",
""
]
] |
2508.14275
|
Cliff O'Reilly
|
Cliff O'Reilly, Ernesto Jimenez-Ruiz, Tillman Weyde
|
Disentangling concept semantics via multilingual averaging in Sparse Autoencoders
| null | null | null | null |
cs.CL cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Connecting LLMs with formal knowledge representation and reasoning is a promising approach to address their shortcomings. Embeddings and sparse autoencoders are widely used to represent textual content, but the semantics are entangled with syntactic and language-specific information. We propose a method that isolates concept semantics in Large Langue Models by averaging concept activations derived via Sparse Autoencoders. We create English text representations from OWL ontology classes, translate the English into French and Chinese and then pass these texts as prompts to the Gemma 2B LLM. Using the open source Gemma Scope suite of Sparse Autoencoders, we obtain concept activations for each class and language version. We average the different language activations to derive a conceptual average. We then correlate the conceptual averages with a ground truth mapping between ontology classes. Our results give a strong indication that the conceptual average aligns to the true relationship between classes when compared with a single language by itself. The result hints at a new technique which enables mechanistic interpretation of internal network states with higher accuracy.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:18:56 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"O'Reilly",
"Cliff",
""
],
[
"Jimenez-Ruiz",
"Ernesto",
""
],
[
"Weyde",
"Tillman",
""
]
] |
2508.14276
|
Mostafa Mehdipour Ghazi
|
Said Djafar Said, Torkan Gholamalizadeh, Mostafa Mehdipour Ghazi
|
Tooth-Diffusion: Guided 3D CBCT Synthesis with Fine-Grained Tooth Conditioning
|
MICCAI 2025 Workshop on Oral and Dental Image Analysis (ODIN)
| null | null | null |
cs.CV cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Despite the growing importance of dental CBCT scans for diagnosis and treatment planning, generating anatomically realistic scans with fine-grained control remains a challenge in medical image synthesis. In this work, we propose a novel conditional diffusion framework for 3D dental volume generation, guided by tooth-level binary attributes that allow precise control over tooth presence and configuration. Our approach integrates wavelet-based denoising diffusion, FiLM conditioning, and masked loss functions to focus learning on relevant anatomical structures. We evaluate the model across diverse tasks, such as tooth addition, removal, and full dentition synthesis, using both paired and distributional similarity metrics. Results show strong fidelity and generalization with low FID scores, robust inpainting performance, and SSIM values above 0.91 even on unseen scans. By enabling realistic, localized modification of dentition without rescanning, this work opens opportunities for surgical planning, patient communication, and targeted data augmentation in dental AI workflows. The codes are available at: https://github.com/djafar1/tooth-diffusion.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:21:35 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Said",
"Said Djafar",
""
],
[
"Gholamalizadeh",
"Torkan",
""
],
[
"Ghazi",
"Mostafa Mehdipour",
""
]
] |
2508.14277
|
Sadmir Kudin
|
Sadmir Kudin, Enes Pasalic, Alexandr Polujan and Fengrong Zhang
|
Permutations satisfying $(P_1)$ and $(P_2)$ properties and $\ell$-optimal bent functions
| null | null | null | null |
math.CO
|
http://creativecommons.org/licenses/by/4.0/
|
An important classification of permutations over $\mathbb{F}_2^m$, suitable for constructing Maiorana-McFarland bent functions on $\mathbb{F}_2^m \times \mathbb{F}_2^m$ with the unique $M$-subspace of maximal dimension, was recently considered in Pasalic et al. (IEEE Trans. Inf. Theory 70(6): 4464-4477, 2024). More precisely, two properties called $(P_1)$ and $(P_2)$ were introduced and a generic method of constructing permutations having the property $(P_1)$ was presented, whereas no such results were provided related to the $(P_2)$ property. In this article, we provide a deeper insight on these properties, their mutual relationship, and specify some explicit classes of permutations having these properties. Such permutations are then employed to generate a large variety of bent functions outside the completed Maiorana-McFarland class $M^\#$. We also introduce $\ell$-optimal bent functions as bent functions with the lowest possible linearity index; such functions can be considered as opposite to Maiorana-McFarland bent functions. We give explicit constructions of $\ell$-optimal bent functions within the $D_0$ class, which in turn can be employed in certain secondary constructions of bent functions for providing even more classes of bent functions that are provably outside $M^\#$. Moreover, we demonstrate that a certain subclass of $D_0$ has an additional property of having only 5-valued spectra decompositions. Finally, we generalize the so-called "swapping variables" method which then allows us to specify large families of bent functions outside $M^\#$. In this way, we give a better explanation of the origin of bent functions in dimension eight, since the vast majority of them is outside ${M}^\#$.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:23:50 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Kudin",
"Sadmir",
""
],
[
"Pasalic",
"Enes",
""
],
[
"Polujan",
"Alexandr",
""
],
[
"Zhang",
"Fengrong",
""
]
] |
2508.14279
|
Adrian Marius Dumitran
|
Adrian-Marius Dumitran and Alexandra-Mihaela Danila and Angela-Liliana Dumitran
|
GRILE: A Benchmark for Grammar Reasoning and Explanation in Romanian LLMs
|
Accepted as long paper @RANLP2025
| null | null | null |
cs.CL cs.CY
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
LLMs (Large language models) have revolutionized NLP (Natural Language Processing), yet their pedagogical value for low-resource languages remains unclear. We present GRILE (Grammar Romanian Inference and Language Explanations) , the first open benchmark of 1,151 multiple-choice questions harvested from Romanian high-stakes exams (National Evaluation, Baccalaureate, university admissions). GRILE enables us to probe two complementary abilities of seven state-of-the-art multilingual and Romanian-specific LLMs: (i) selecting the correct answer, and (ii) producing linguistically accurate explanations. While Gemini 2.5 Pro reaches 83% accuracy, most open-weight models stay below 65%, and 48% of their explanations contain factual or pedagogical flaws according to expert review. A detailed error analysis pinpoints systematic weaknesses in morphology and in applying the latest DOOM3 orthographic norms. All data, code and a public web demo are released to catalyze future research. Our findings expose open challenges for trustworthy educational NLP in low-resource settings and establish GRILE as a new test-bed for controllable explanation generation and evaluation.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:27:06 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Dumitran",
"Adrian-Marius",
""
],
[
"Danila",
"Alexandra-Mihaela",
""
],
[
"Dumitran",
"Angela-Liliana",
""
]
] |
2508.14280
|
Ali Rasekh
|
Ali Rasekh, Sepehr Kazemi Ranjbar, Simon Gottschalk
|
Multi-Rationale Explainable Object Recognition via Contrastive Conditional Inference
| null | null | null | null |
cs.CV
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Explainable object recognition using vision-language models such as CLIP involves predicting accurate category labels supported by rationales that justify the decision-making process. Existing methods typically rely on prompt-based conditioning, which suffers from limitations in CLIP's text encoder and provides weak conditioning on explanatory structures. Additionally, prior datasets are often restricted to single, and frequently noisy, rationales that fail to capture the full diversity of discriminative image features. In this work, we introduce a multi-rationale explainable object recognition benchmark comprising datasets in which each image is annotated with multiple ground-truth rationales, along with evaluation metrics designed to offer a more comprehensive representation of the task. To overcome the limitations of previous approaches, we propose a contrastive conditional inference (CCI) framework that explicitly models the probabilistic relationships among image embeddings, category labels, and rationales. Without requiring any training, our framework enables more effective conditioning on rationales to predict accurate object categories. Our approach achieves state-of-the-art results on the multi-rationale explainable object recognition benchmark, including strong zero-shot performance, and sets a new standard for both classification accuracy and rationale quality. Together with the benchmark, this work provides a more complete framework for evaluating future models in explainable object recognition. The code will be made available online.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:28:12 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Rasekh",
"Ali",
""
],
[
"Ranjbar",
"Sepehr Kazemi",
""
],
[
"Gottschalk",
"Simon",
""
]
] |
2508.14281
|
Zhun Yin
|
Zhun Yin, Xiaotian Li, Lifan Mei, Yong Liu, Zhong-Ping Jiang
|
DeeP-TE: Data-enabled Predictive Traffic Engineering
| null | null | null | null |
cs.NI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Routing configurations of a network should constantly adapt to traffic variations to achieve good network performance. Adaptive routing faces two main challenges: 1) how to accurately measure/estimate time-varying traffic matrices? 2) how to control the network and application performance degradation caused by frequent route changes? In this paper, we develop a novel data-enabled predictive traffic engineering (DeeP-TE) algorithm that minimizes the network congestion by gracefully adapting routing configurations over time. Our control algorithm can generate routing updates directly from the historical routing data and the corresponding link rate data, without direct traffic matrix measurement or estimation. Numerical experiments on real network topologies with real traffic matrices demonstrate that the proposed DeeP-TE routing adaptation algorithm can achieve close-to-optimal control effectiveness with significantly lower routing variations than the baseline methods.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:37:25 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Yin",
"Zhun",
""
],
[
"Li",
"Xiaotian",
""
],
[
"Mei",
"Lifan",
""
],
[
"Liu",
"Yong",
""
],
[
"Jiang",
"Zhong-Ping",
""
]
] |
2508.14283
|
Liangtao Peng
|
Christophe De Beule, Liangtao Peng, E. J. Mele, and Shaffique Adam
|
Role of electron-electron interactions in $M$-valley twisted transition metal dichalcogenides
| null | null | null | null |
cond-mat.mes-hall cond-mat.str-el
|
http://creativecommons.org/licenses/by/4.0/
|
We investigate the role of long-range Coulomb interactions in $M$-valley moir\'e systems using the self-consistent Hartree-Fock approximation. This platform was recently proposed [Nature 643, 376 (2025) and arXiv:2411.18828 (2024)] as a new class of experimentally realizable moir\'e materials using twisted transition metal dichalcogenides homobilayers with the 1T structure. While these seminal studies considered the noninteracting theory without an interlayer bias due to an electric displacement field, this work shows that both electron-electron interactions at finite doping and an interlayer bias strongly modify the moir\'e bands. For small twist angles, the density of states as a function of filling and interlayer bias shows qualitatively different behavior for twisting near aligned ($0^\circ$) and antialigned ($60^\circ$) stacking. More interestingly, the Van Hove singularity becomes pinned to the Fermi energy over a finite range of doping, an effect known to enhance both superconductivity and strongly correlated states. For aligned stacking this occurs only at zero electric field, while for antialigned stacking this happens both at zero and finite field. Our work demonstrates that correlated states in $M$-valley 1TtTMDs can be strongly tuned \textit{in situ} both by applying an electric displacement field and by electron doping.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:43:46 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"De Beule",
"Christophe",
""
],
[
"Peng",
"Liangtao",
""
],
[
"Mele",
"E. J.",
""
],
[
"Adam",
"Shaffique",
""
]
] |
2508.14284
|
Jonathan Passerat-Palmbach
|
Jonathan Passerat-Palmbach and Sarisht Wadhwa
|
Differentially Private aggregate hints in mev-share
| null | null | null | null |
cs.CR
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Flashbots recently released mev-share to empower users with control over the amount of information they share with searchers for extracting Maximal Extractable Value (MEV). Searchers require more information to maintain on-chain exchange efficiency and profitability, while users aim to prevent frontrunning by withholding information. After analyzing two searching strategies in mev-share to reason about searching techniques, this paper introduces Differentially-Private (DP) aggregate hints as a new type of hints to disclose information quantitatively. DP aggregate hints enable users to formally quantify their privacy loss to searchers, and thus better estimate the level of rebates to ask in return. The paper discusses the current properties and privacy loss in mev-share and lays out how DP aggregate hints could enhance the system for both users and searchers. We leverage Differential Privacy in the Trusted Curator Model to design our aggregate hints. Additionally, we explain how random sampling can defend against sybil attacks and amplify overall user privacy while providing valuable hints to searchers for improved backrunning extraction and frontrunning prevention.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:53:31 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Passerat-Palmbach",
"Jonathan",
""
],
[
"Wadhwa",
"Sarisht",
""
]
] |
2508.14285
|
Liyi Zhang
|
Liyi Zhang, Jake Snell, Thomas L. Griffiths
|
Amortized Bayesian Meta-Learning for Low-Rank Adaptation of Large Language Models
|
6 pages, 2 figures
| null | null | null |
cs.LG cs.AI stat.ML
|
http://creativecommons.org/licenses/by/4.0/
|
Fine-tuning large language models (LLMs) with low-rank adaptaion (LoRA) is a cost-effective way to incorporate information from a specific dataset. However, it is often unclear how well the fine-tuned LLM will generalize, i.e., how well it will perform on unseen datasets. Methods have been proposed to improve generalization by optimizing with in-context prompts, or by using meta-learning to fine-tune LLMs. However, these methods are expensive in memory and computation, requiring either long-context prompts or saving copies of parameters and using second-order gradient updates. To address these challenges, we propose Amortized Bayesian Meta-Learning for LoRA (ABMLL). This method builds on amortized Bayesian meta-learning for smaller models, adapting this approach to LLMs while maintaining its computational efficiency. We reframe task-specific and global parameters in the context of LoRA and use a set of new hyperparameters to balance reconstruction accuracy and the fidelity of task-specific parameters to the global ones. ABMLL provides effective generalization and scales to large models such as Llama3-8B. Furthermore, as a result of using a Bayesian framework, ABMLL provides improved uncertainty quantification. We test ABMLL on Unified-QA and CrossFit datasets and find that it outperforms existing methods on these benchmarks in terms of both accuracy and expected calibration error.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:57:59 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zhang",
"Liyi",
""
],
[
"Snell",
"Jake",
""
],
[
"Griffiths",
"Thomas L.",
""
]
] |
2508.14286
|
Anushka Ashish Kore
|
Anushka A. Kore, Frank G. te Nijenhuis, Matthijs van der Sluijs, Wim van Zwam, Charles Majoie, Geert Lycklama \`a Nijeholt, Danny Ruijters, Frans Vos, Sandra Cornelissen, Ruisheng Su, Theo van Walsum
|
OccluNet: Spatio-Temporal Deep Learning for Occlusion Detection on DSA
|
To be published in Proceedings of the SWITCH Workshop at MICCAI 2025, Lecture Notes in Computer Science (LNCS), Springer
| null | null | null |
cs.CV cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Accurate detection of vascular occlusions during endovascular thrombectomy (EVT) is critical in acute ischemic stroke (AIS). Interpretation of digital subtraction angiography (DSA) sequences poses challenges due to anatomical complexity and time constraints. This work proposes OccluNet, a spatio-temporal deep learning model that integrates YOLOX, a single-stage object detector, with transformer-based temporal attention mechanisms to automate occlusion detection in DSA sequences. We compared OccluNet with a YOLOv11 baseline trained on either individual DSA frames or minimum intensity projections. Two spatio-temporal variants were explored for OccluNet: pure temporal attention and divided space-time attention. Evaluation on DSA images from the MR CLEAN Registry revealed the model's capability to capture temporally consistent features, achieving precision and recall of 89.02% and 74.87%, respectively. OccluNet significantly outperformed the baseline models, and both attention variants attained similar performance. Source code is available at https://github.com/anushka-kore/OccluNet.git
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 21:59:59 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Kore",
"Anushka A.",
""
],
[
"Nijenhuis",
"Frank G. te",
""
],
[
"van der Sluijs",
"Matthijs",
""
],
[
"van Zwam",
"Wim",
""
],
[
"Majoie",
"Charles",
""
],
[
"Nijeholt",
"Geert Lycklama à",
""
],
[
"Ruijters",
"Danny",
""
],
[
"Vos",
"Frans",
""
],
[
"Cornelissen",
"Sandra",
""
],
[
"Su",
"Ruisheng",
""
],
[
"van Walsum",
"Theo",
""
]
] |
2508.14287
|
Or Vardi
|
Yossi Azar, Debmalya Panigrahi, Or Vardi
|
Nearly Tight Bounds for the Online Sorting Problem
| null | null | null | null |
cs.DS
|
http://creativecommons.org/licenses/by/4.0/
|
In the online sorting problem, a sequence of $n$ numbers in $[0, 1]$ (including $\{0,1\}$) have to be inserted in an array of size $m \ge n$ so as to minimize the sum of absolute differences between pairs of numbers occupying consecutive non-empty cells. Previously, Aamand {\em et al.} (SODA 2023) gave a deterministic $2^{\sqrt{\log n} \sqrt{\log \log n + \log (1/\varepsilon)}}$-competitive algorithm when $m = (1+\varepsilon) n$ for any $\varepsilon \ge \Omega(\log n/n)$. They also showed a lower bound: with $m = \gamma n$ space, the competitive ratio of any deterministic algorithm is at least $\frac{1}{\gamma}\cdot\Omega(\log n / \log \log n)$. This left an exponential gap between the upper and lower bounds for the problem.
In this paper, we bridge this exponential gap and almost completely resolve the online sorting problem. First, we give a deterministic $O(\log^2 n / \varepsilon)$-competitive algorithm with $m = (1+\varepsilon) n$, for any $\varepsilon \ge \Omega(\log n / n)$. Next, for $m = \gamma n$ where $\gamma = [O(1), O(\log^2 n)]$, we give a deterministic $O(\log^2 n / \gamma)$-competitive algorithm. In particular, this implies an $O(1)$-competitive algorithm with $O(n \log^2 n)$ space, which is within an $O(\log n\cdot \log \log n)$ factor of the lower bound of $\Omega(n \log n / \log \log n)$. Combined, the two results imply a close to optimal tradeoff between space and competitive ratio for the entire range of interest: specifically, an upper bound of $O(\log^2 n)$ on the product of the competitive ratio and $\gamma$ while the lower bound on this product is $\Omega(\log n / \log\log n)$. We also show that these results can be extended to the case when the range of the numbers is not known in advance, for an additional $O(\log n)$ factor in the competitive ratio.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:05:21 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Azar",
"Yossi",
""
],
[
"Panigrahi",
"Debmalya",
""
],
[
"Vardi",
"Or",
""
]
] |
2508.14288
|
Yewei Song
|
Yewei Song, Tiezhu Sun, Xunzhu Tang, Prateek Rajput, Tegawende F. Bissyande, Jacques Klein
|
Measuring LLM Code Generation Stability via Structural Entropy
|
ASE-NIER
| null | null | null |
cs.SE cs.CL
|
http://creativecommons.org/licenses/by-sa/4.0/
|
Assessing the stability of code generation from large language models (LLMs) is essential for judging their reliability in real-world development. We extend prior "structural-entropy concepts" to the program domain by pairing entropy with abstract syntax tree (AST) analysis. For any fixed prompt, we collect the multiset of depth-bounded subtrees of AST in each generated program and treat their relative frequencies as a probability distribution. We then measure stability in two complementary ways: (i) Jensen-Shannon divergence, a symmetric, bounded indicator of structural overlap, and (ii) a Structural Cross-Entropy ratio that highlights missing high-probability patterns. Both metrics admit structural-only and token-aware variants, enabling separate views on control-flow shape and identifier-level variability. Unlike pass@k, BLEU, or CodeBLEU, our metrics are reference-free, language-agnostic, and execution-independent. We benchmark several leading LLMs on standard code generation tasks, demonstrating that AST-driven structural entropy reveals nuances in model consistency and robustness. The method runs in O(n,d) time with no external tests, providing a lightweight addition to the code-generation evaluation toolkit.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:07:12 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Song",
"Yewei",
""
],
[
"Sun",
"Tiezhu",
""
],
[
"Tang",
"Xunzhu",
""
],
[
"Rajput",
"Prateek",
""
],
[
"Bissyande",
"Tegawende F.",
""
],
[
"Klein",
"Jacques",
""
]
] |
2508.14289
|
Dylan Cashman
|
Areen Khalaila, Lane Harrison, Nam Wook Kim, Dylan Cashman
|
"They Aren't Built For Me": An Exploratory Study of Strategies for Measurement of Graphical Primitives in Tactile Graphics
| null | null | null | null |
cs.HC
|
http://creativecommons.org/licenses/by/4.0/
|
Advancements in accessibility technologies such as low-cost swell form printers or refreshable tactile displays promise to allow blind or low-vision (BLV) people to analyze data by transforming visual representations directly to tactile representations. However, it is possible that design guidelines derived from experiments on the visual perception system may not be suited for the tactile perception system. We investigate the potential mismatch between familiar visual encodings and tactile perception in an exploratory study into the strategies employed by BLV people to measure common graphical primitives converted to tactile representations. First, we replicate the Cleveland and McGill study on graphical perception using swell form printing with eleven BLV subjects. Then, we present results from a group interview in which we describe the strategies used by our subjects to read four common chart types. While our results suggest that familiar encodings based on visual perception studies can be useful in tactile graphics, our subjects also expressed a desire to use encodings designed explicitly for BLV people. Based on this study, we identify gaps between the perceptual expectations of common charts and the perceptual tools available in tactile perception. Then, we present a set of guidelines for the design of tactile graphics that accounts for these gaps. Supplemental material is available at https://osf.io/3nsfp/?view_only=7b7b8dcbae1d4c9a8bb4325053d13d9f.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:11:40 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Khalaila",
"Areen",
""
],
[
"Harrison",
"Lane",
""
],
[
"Kim",
"Nam Wook",
""
],
[
"Cashman",
"Dylan",
""
]
] |
2508.14290
|
Pace Nielsen
|
Pace P. Nielsen
|
Algebraization of infinite summation
|
47 pages, comments welcome
| null | null | null |
math.RA math.AC math.CA math.CT math.GN
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
An algebraic framework in which to study infinite sums is proposed, complementing and augmenting the usual topological tools. The framework subsumes numerous examples in the literature. It is developed using many varied examples, with a particular emphasis on infinitizing the usual group and ring axioms. Comparing these examples reveals that a few key algebraic properties play a crucial role in the behaviors of different forms of infinite summation. Special attention is given to associativity, which is particularly difficult to properly infinitize. In that context, there is an important technique called the Eilenberg-Mazer swindle that is studied and greatly generalized.
Some special properties are singled out as potential axioms. Interactions between these potential axioms are analyzed, and numerous results explore how to impose new axioms while retaining old ones. In some cases the axioms classify or categorize a given example. Surprisingly, such axiomatizations are very concise, relying on only a handful of natural conditions.
These investigations reveal more precisely the part that topology plays in the formation of infinite sums. Special attention is given to the methods of partial summation and unconditional summation. In the opposite direction, it is proved that from the infinite sums alone one can create a refined topology, lying between the original topology and its sequential coreflection. Another especially interesting application of these ideas is the construction of new algebraic quotient structures that retain the ability to handle infinite summation.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:13:39 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Nielsen",
"Pace P.",
""
]
] |
2508.14291
|
Martin Frankland
|
Martin Frankland, Arnaud Ngopnang Ngomp\'e
|
Enriched model categories and the Dold-Kan correspondence
| null | null | null | null |
math.AT math.CT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
The monoidal properties of the Dold-Kan correspondence have been studied in homotopy theory, notably by Schwede and Shipley. Changing the enrichment of an enriched, tensored, and cotensored category along the Dold-Kan correspondence does not preserve the tensoring nor the cotensoring. More generally, what happens to an enriched model category if we change the enrichment along a weak monoidal Quillen pair? We prove a change of base theorem that describes which properties are preserved and which are weakened.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:14:30 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Frankland",
"Martin",
""
],
[
"Ngompé",
"Arnaud Ngopnang",
""
]
] |
2508.14292
|
M. Ali Bayram
|
M. Ali Bayram, Ali Arda Fincan, Ahmet Semih G\"um\"u\c{s}, Sercan Karaka\c{s}, Banu Diri, Sava\c{s} Y{\i}ld{\i}r{\i}m, Demircan \c{C}elik
|
Tokens with Meaning: A Hybrid Tokenization Approach for NLP
| null | null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
Tokenization plays a pivotal role in natural language processing (NLP), shaping how text is segmented and interpreted by language models. While subword methods such as Byte Pair Encoding (BPE) and WordPiece have been effective, they often struggle with morphologically rich and agglutinative languages because they rely on frequency rather than linguistic structure. We introduce a hybrid tokenization framework that combines rule-based morphological analysis with statistical subword segmentation. The method uses phonological normalization, root-affix dictionaries, and a novel algorithm that balances morpheme preservation with vocabulary efficiency. It assigns shared identifiers to phonologically variant affixes (e.g., -ler and -lar) and altered root forms (e.g., kitap vs. kitab{\i}), reducing redundancy while maintaining semantic integrity. Special tokens are added for whitespace and case, including an UPPERCASE marker to avoid vocabulary inflation from capitalization. BPE is integrated for out-of-vocabulary coverage without harming morphological coherence. On the TR-MMLU benchmark, the tokenizer achieves the highest Turkish Token Percentage (90.29\%) and Pure Token Percentage (85.8\%). Comparisons with tokenizers from LLaMA, Gemma, and GPT show more linguistically meaningful and coherent tokens. Although demonstrated on Turkish, the approach is language-independent and adaptable to other languages, offering a practical path toward more interpretable and effective multilingual NLP systems.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:17:42 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Bayram",
"M. Ali",
""
],
[
"Fincan",
"Ali Arda",
""
],
[
"Gümüş",
"Ahmet Semih",
""
],
[
"Karakaş",
"Sercan",
""
],
[
"Diri",
"Banu",
""
],
[
"Yıldırım",
"Savaş",
""
],
[
"Çelik",
"Demircan",
""
]
] |
2508.14293
|
Jason Gombas
|
Jason Peter Gombas
|
Search for a Heavy-philic W' Boson using Proton-Proton Collisions at Center-of-Mass Energy of 13 TeV Using the ATLAS Detector
|
PhD thesis
| null | null | null |
hep-ex
|
http://creativecommons.org/licenses/by/4.0/
|
This thesis presents a search for a new, hypothetical particle predicted by theories extending the Standard Model of particle physics. This heavy $W'$ boson interacts only with the heaviest known quarks, top and bottom (heavy-philic). Such a particle could provide insight into the fundamental forces of nature and be the first hint at extra dimensions or a composite Higgs. The $W'$ boson is produced in high-energy proton-proton collisions, mainly through gluon fusion. Its decay leads to a distinctive final state: $tbW'\rightarrow tbtb$. This search uses data collected by the ATLAS detector during Run 2 at the Large Hadron Collider, focusing on events with a single charged lepton, at least five jets, and at least three jets identified as originating from bottom quarks. To improve sensitivity to this rare process, advanced machine learning techniques are applied. A profile likelihood fit to the machine learning output is used to evaluate the data. No significant excess above the Standard Model background is observed, and exclusion limits are set at the 95\% confidence level on the production cross-section of the heavy-philic $W'$ boson.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:19:04 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Gombas",
"Jason Peter",
""
]
] |
2508.14294
|
Maria Leonor Pacheco
|
Maria Leonor Pacheco, Fabio Somenzi, Dananjay Srinivas, Ashutosh Trivedi
|
Explaining Hitori Puzzles: Neurosymbolic Proof Staging for Sequential Decisions
| null | null | null | null |
cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
We propose a neurosymbolic approach to the explanation of complex sequences of decisions that combines the strengths of decision procedures and Large Language Models (LLMs). We demonstrate this approach by producing explanations for the solutions of Hitori puzzles. The rules of Hitori include local constraints that are effectively explained by short resolution proofs. However, they also include a connectivity constraint that is more suitable for visual explanations. Hence, Hitori provides an excellent testing ground for a flexible combination of SAT solvers and LLMs. We have implemented a tool that assists humans in solving Hitori puzzles, and we present experimental evidence of its effectiveness.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:21:44 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Pacheco",
"Maria Leonor",
""
],
[
"Somenzi",
"Fabio",
""
],
[
"Srinivas",
"Dananjay",
""
],
[
"Trivedi",
"Ashutosh",
""
]
] |
2508.14295
|
Jonathan Hunt
|
Yuguang Yue, Chris Green, Samuel Hunt, Irakli Salia, Wenzhe Shi, Jonathan J Hunt
|
Pixels to Play: A Foundation Model for 3D Gameplay
| null |
Conference on Games 2025 (Short paper)
| null | null |
cs.CV cs.AI cs.LG
|
http://creativecommons.org/licenses/by-sa/4.0/
|
We introduce Pixels2Play-0.1 (P2P0.1), a foundation model that learns to play a wide range of 3D video games with recognizable human-like behavior. Motivated by emerging consumer and developer use cases - AI teammates, controllable NPCs, personalized live-streamers, assistive testers - we argue that an agent must rely on the same pixel stream available to players and generalize to new titles with minimal game-specific engineering. P2P0.1 is trained end-to-end with behavior cloning: labeled demonstrations collected from instrumented human game-play are complemented by unlabeled public videos, to which we impute actions via an inverse-dynamics model. A decoder-only transformer with auto-regressive action output handles the large action space while remaining latency-friendly on a single consumer GPU. We report qualitative results showing competent play across simple Roblox and classic MS-DOS titles, ablations on unlabeled data, and outline the scaling and evaluation steps required to reach expert-level, text-conditioned control.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:24:50 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Yue",
"Yuguang",
""
],
[
"Green",
"Chris",
""
],
[
"Hunt",
"Samuel",
""
],
[
"Salia",
"Irakli",
""
],
[
"Shi",
"Wenzhe",
""
],
[
"Hunt",
"Jonathan J",
""
]
] |
2508.14296
|
Dimiter Hadjimichef
|
Jo\~ao Gabriel Galli Gimenez, Dimiter Hadjimichef, Peter O. Hess, Marcelo Netz-Marzola, C\'esar A. Zen Vasconcellos
|
Can Effects of a Generalized Uncertainty Principle Appear in Compact Stars?
| null |
Universe 11 (2025) 1, 5
|
10.3390/universe11010005
| null |
gr-qc
|
http://creativecommons.org/licenses/by-nc-nd/4.0/
|
In the present contribution, a preliminary analysis of the effects of the Generalized Uncertainty Principle (GUP) with a minimum length, in the context of compact stars, is performed. On basis of a deformed Poisson canonical algebra with a parametrized minimum length scale that induces deviations from conventional Quantum Mechanics, fundamental questions involving the consistence, evidences and proofs of this approach as a possible cure for unbounded energy divergence are outlined. The incorporation of GUP effects into semiclassical 2N-dimensional systems is made by means of a time-invariant distortion transformation applied to their non-deformed counterparts. Assuming the quantum hadrodynamics $\sigma-\omega$ approach as a toy-model, due to its simplicity and structured description of neutron stars, we perform a preliminary analysis of GUP effects with a minimum spacetime length on these compact objects. The corresponding results for the equation of state and the mass-radius relation for neutron stars are in tune with recent observations with a maximum mass around $2.5 M_{\odot}$ and radius close to $12$ km. Our results also indicate the smallness of the noncommutative scale.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:28:09 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Gimenez",
"João Gabriel Galli",
""
],
[
"Hadjimichef",
"Dimiter",
""
],
[
"Hess",
"Peter O.",
""
],
[
"Netz-Marzola",
"Marcelo",
""
],
[
"Vasconcellos",
"César A. Zen",
""
]
] |
2508.14297
|
Jesus Silva Rodriguez
|
Jesus Silva-Rodriguez, Tianxia Zhao, Ran Mo, Xingpeng Li
|
Grid-Edge Energy-Flexible Technologies: A Comparative Analysis Across Generators, Loads, and Energy Storage Systems
| null | null | null | null |
eess.SY cs.SY
|
http://creativecommons.org/licenses/by/4.0/
|
This review analysis presents a comprehensive exploration of energy flexibility in modern power systems. It examines the roles and mechanisms of flexible technologies across three main categories: generators, energy storage systems (ESS), and loads. Energy flexibility is defined as the ability to dynamically adjust supply and/or demand in response to grid conditions to maintain balance and stability. This is of particular importance to facilitate the integration of the growing variable renewable energy sources (RES) into modern power grids. Additionally, traditional supply-side mechanisms to maintain balance and stability are complemented by advancements in demand-side management and demand response strategies, which enable loads to adjust consumption patterns and schedules in response to grid requirements. ESS are also explored to further enhance flexibility by absorbing excess generation and/or supplying large load increases that are not able to be met by the less flexible resources. This paper also explores specific flexibility technologies, examining their characteristics, control strategies, advantages, and limitations. Energy flexibility services are also categorized into intermittency mitigation, peak shaving, and energy reserve provisioning. Each service is supported by case studies and examples demonstrating how different resources respond to varying conditions. Ultimately, the findings and reviews of the various flexible resources in this paper provide a roadmap for optimizing energy flexibility across diverse resource types, paving the way for a more sustainable and resilient energy future.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:32:13 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Silva-Rodriguez",
"Jesus",
""
],
[
"Zhao",
"Tianxia",
""
],
[
"Mo",
"Ran",
""
],
[
"Li",
"Xingpeng",
""
]
] |
2508.14298
|
Andrei Belitsky
|
A.V. Belitsky, V.A. Smirnov
|
Tropical regions of near mass-shell pentabox
|
37 pages, 5 figures, 1 Mathematica notebook, and multiple ancillary files
| null | null | null |
hep-th hep-ph
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Coulomb branch amplitudes of maximally supersymmetric Yang-Mills theory display infrared properties different from their conformal counterparts. While the four-leg amplitude is known to very high perturbative orders, amplitudes of higher multiplicity fall into an uncharted territory starting already from two loops. The reason for this is that they are not easily amenable to traditional techniques like canonical differential equations due to the uncontrolled swelling of solutions to integration-by-parts identities. In this paper, we break the barrier for the five-leg amplitude using a technique based on the analysis of Newton polytopes corresponding to Feynman/Schwinger integrands and their tropical geometry. Specifically, we analytically evaluate the near mass-shell limit of the off-shell pentabox in terms of Goncharov polylogarithms.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:32:26 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Belitsky",
"A. V.",
""
],
[
"Smirnov",
"V. A.",
""
]
] |
2508.14299
|
Minsen Yuan
|
Minsen Yuan and Yue Yu
|
Sequential Convex Programming with Filtering-Based Warm-Starting for Continuous-Time Multiagent Quadrotor Trajectory Optimization
| null | null | null | null |
math.OC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Optimizing the trajectories of multiple quadrotors in a shared space is a core challenge in various applications. Many existing trajectory optimization methods enforce constraints only at the discretization points, leading to violations between discretization points. They also often lack warm-starting strategies for iterative solution methods such as sequential convex programming, causing slow convergence or sensitivity to the initial guess. We propose a framework for optimizing multiagent quadrotor trajectories that combines a sequential convex programming approach with filtering-based warm-starting. This framework not only ensures constraint satisfaction along the entire continuous-time trajectory but also provides an online warm-starting strategy that accelerates convergence and improves solution quality in numerical experiments. The key idea is to first transform continuous-time constraints into auxiliary nonlinear dynamics and boundary constraints, both of which are compatible with sequential convex programming. Furthermore, we propose a novel warm-starting strategy by approximating the trajectory optimization problem as a Bayesian state estimation problem. This approximation provides an efficient estimate of the optimal trajectories. We demonstrate the proposed framework on multiagent quadrotor trajectory optimization with collision avoidance constraints. Compared with benchmark methods, the proposed framework achieves not only continuous-time constraint satisfaction, but also reduces computation time by up to two orders of magnitude.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:40:29 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Yuan",
"Minsen",
""
],
[
"Yu",
"Yue",
""
]
] |
2508.14300
|
Youssef Maklad
|
Youssef Maklad, Fares Wael, Ali Hamdi, Wael Elsersy, Khaled Shaban
|
MultiFuzz: A Dense Retrieval-based Multi-Agent System for Network Protocol Fuzzing
| null | null | null | null |
cs.CR cs.CL cs.MA cs.NI
|
http://creativecommons.org/licenses/by/4.0/
|
Traditional protocol fuzzing techniques, such as those employed by AFL-based systems, often lack effectiveness due to a limited semantic understanding of complex protocol grammars and rigid seed mutation strategies. Recent works, such as ChatAFL, have integrated Large Language Models (LLMs) to guide protocol fuzzing and address these limitations, pushing protocol fuzzers to wider exploration of the protocol state space. But ChatAFL still faces issues like unreliable output, LLM hallucinations, and assumptions of LLM knowledge about protocol specifications. This paper introduces MultiFuzz, a novel dense retrieval-based multi-agent system designed to overcome these limitations by integrating semantic-aware context retrieval, specialized agents, and structured tool-assisted reasoning. MultiFuzz utilizes agentic chunks of protocol documentation (RFC Documents) to build embeddings in a vector database for a retrieval-augmented generation (RAG) pipeline, enabling agents to generate more reliable and structured outputs, enhancing the fuzzer in mutating protocol messages with enhanced state coverage and adherence to syntactic constraints. The framework decomposes the fuzzing process into modular groups of agents that collaborate through chain-of-thought reasoning to dynamically adapt fuzzing strategies based on the retrieved contextual knowledge. Experimental evaluations on the Real-Time Streaming Protocol (RTSP) demonstrate that MultiFuzz significantly improves branch coverage and explores deeper protocol states and transitions over state-of-the-art (SOTA) fuzzers such as NSFuzz, AFLNet, and ChatAFL. By combining dense retrieval, agentic coordination, and language model reasoning, MultiFuzz establishes a new paradigm in autonomous protocol fuzzing, offering a scalable and extensible foundation for future research in intelligent agentic-based fuzzing systems.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:42:04 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Maklad",
"Youssef",
""
],
[
"Wael",
"Fares",
""
],
[
"Hamdi",
"Ali",
""
],
[
"Elsersy",
"Wael",
""
],
[
"Shaban",
"Khaled",
""
]
] |
2508.14301
|
Douglas Cumming
|
Douglas Cumming
|
The Value of Finance Journals
|
24 pages, 5 tables, 2 figures
| null | null | null |
econ.GN q-fin.EC
|
http://creativecommons.org/licenses/by/4.0/
|
I present data from SCImago, 1999-2024, on factors that affect the valuation of finance journals. The SCImago data show some interesting facts; for example, in 2024, Finance Research Letters, valued at \$16.5 million, was four times more valuable than the Journal of Financial Economics (\$4.1m), five times more valuable than the Review of Financial Studies (\$3.3m), and over seven times more valuable than the Journal of Finance (\$2.3m). The data show that quality, as proxied by citations, positively affects journal value, and this value increase is heightened by citations from policy documents; similarly, the absence of quality, or a greater percentage of uncited documents, negatively affects journal value. The data indicate that the creation of the Elsevier Ecosystem of networked finance journals significantly enhances journal value. Further, a greater percentage of female authors positively impacts value. I discuss policy implications and suggestions for future work based on the evidence presented.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:50:03 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Cumming",
"Douglas",
""
]
] |
2508.14302
|
Negar Hassanpour
|
Amirmohsen Sattarifard, Sepehr Lavasani, Ehsan Imani, Kunlin Zhang, Hanlin Xu, Fengyu Sun, Negar Hassanpour, Chao Gao
|
GLASS: Test-Time Acceleration for LLMs via Global-Local Neural Importance Aggregation
| null | null | null | null |
cs.LG cs.AI cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
Deploying Large Language Models (LLMs) on edge hardware demands aggressive, prompt-aware dynamic pruning to reduce computation without degrading quality. Static or predictor-based schemes either lock in a single sparsity pattern or incur extra runtime overhead, and recent zero-shot methods that rely on statistics from a single prompt fail on short prompt and/or long generation scenarios. We introduce A/I-GLASS: Activation- and Impact-based Global-Local neural importance Aggregation for feed-forward network SparSification, two training-free methods that dynamically select FFN units using a rank-aggregation of prompt local and model-intrinsic global neuron statistics. Empirical results across multiple LLMs and benchmarks demonstrate that GLASS significantly outperforms prior training-free methods, particularly in challenging long-form generation scenarios, without relying on auxiliary predictors or adding any inference overhead.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:50:20 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Sattarifard",
"Amirmohsen",
""
],
[
"Lavasani",
"Sepehr",
""
],
[
"Imani",
"Ehsan",
""
],
[
"Zhang",
"Kunlin",
""
],
[
"Xu",
"Hanlin",
""
],
[
"Sun",
"Fengyu",
""
],
[
"Hassanpour",
"Negar",
""
],
[
"Gao",
"Chao",
""
]
] |
2508.14303
|
Xiaodong An
|
Xiaodong An, Mikael Toye, Flavio H. Fenton
|
Quantifying The Complex Spatiotemporal Chaos of Cardiac Fibrillation in Ionic Models Across Parameter Regimes
| null | null | null | null |
nlin.CD
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Quantifying the complexity of cardiac systems is fundamental to understanding the onset of rhythm disorders, from mild arrhythmias to life-threatening fibrillation. In this work, we investigate how chaos shows up and evolves in simplified cardiac models by calculating the Lyapunov exponent (LE) across different parameter sets. We show that both temporal and spatial LE estimators can be effectively applied to action potential duration (APD) data, even without full access to state variables. Specifically, the spatial-temporal algorithm and Wolf's algorithm are used in quantifying the complexity, with experiments demonstrating their distinct behaviors on various single-spiral patterns. We also identify the minimum data length and sampling density necessary to achieve robust and accurate estimation. Overall, our results suggest that these APD-based methods can be applied not only to simulation data but also to future clinical or experimental data, particularly when observations are limited, such as when only APD data are available for analysis.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:52:19 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"An",
"Xiaodong",
""
],
[
"Toye",
"Mikael",
""
],
[
"Fenton",
"Flavio H.",
""
]
] |
2508.14304
|
Raymond Remigio
|
Raymond P. Remigio, Vivian U, Aaron J. Barth, Nico Winkel, Vardha N. Bennert, Tommaso Treu, Matthew A. Malkan, Sebastian Contreras, Peter R. Williams, Jordan N. Runco, Liam Hunt
|
Spatially Resolved [O III] Emission Line Kinematics of Reverberation-Mapped AGNs with the Keck Cosmic Web Imager
|
28 pages, 13 figures, 4 tables. Accepted for publication in The Astrophysical Journal. The online journal will contain the figure set associated with Figure 3
| null | null | null |
astro-ph.GA
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
We present optical integral-field spectroscopic data of ten nearby ($0.02\leq z \leq 0.05$) Seyfert 1 galaxies taken with the Keck Cosmic Web Imager (KCWI). We map the spatially resolved kinematics of the [O III] gas and stars, and investigate the alignments between their global kinematic position angles (PA). Large-scale gas motions are primarily dominated by rotation, and are kinematically aligned with the stars ($\Delta\text{PA}\leq 30$ deg). However, eight galaxies exhibit non-rotational kinematic signatures (e.g., kinematic twists, possible outflows) in their ionized gas velocity fields near the nucleus. We compare aperture-wide measurements of the gas and stellar velocity dispersions ($\sigma_{\text{gas}}$ and $\sigma_\star$) to test the use of the width of the [O III] line core as a surrogate for $\sigma_\star$. Direct comparisons between $\sigma_{\text{gas}}$ and $\sigma_\star$ show that $\sigma_{\text{gas}}$ tends to underestimate $\sigma_\star$, and thus is not a reliable tracer of $\sigma_\star$ for our selected galaxies. We measure the extent of the narrow-line region (NLR) using several definitions, resulting in sizes of $\sim0.1$-$10$ kpc. For a given [O III] luminosity, our NLR sizes derived from the [O III]/H$\beta$ flux ratio or an [O III] isophotal radius are an order of magnitude larger than those measured from past imaging data.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:58:16 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Remigio",
"Raymond P.",
""
],
[
"U",
"Vivian",
""
],
[
"Barth",
"Aaron J.",
""
],
[
"Winkel",
"Nico",
""
],
[
"Bennert",
"Vardha N.",
""
],
[
"Treu",
"Tommaso",
""
],
[
"Malkan",
"Matthew A.",
""
],
[
"Contreras",
"Sebastian",
""
],
[
"Williams",
"Peter R.",
""
],
[
"Runco",
"Jordan N.",
""
],
[
"Hunt",
"Liam",
""
]
] |
2508.14305
|
Mterorga Ukor
|
Terlumun Gbaden, Mterorga Ukor, Grace Erdoo Ateata
|
Design and Simulation of Fault-Tolerant Network Switching System Using Python-Based Algorithms
|
8 pages, 6 figures
| null | null | null |
cs.NI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Ensuring uninterrupted data flow in modern networks requires robust fault-tolerant mechanisms, especially in environments where reliability and responsiveness are critical. This paper presents the design and simulation of a fault-tolerant network switching system using Python-based algorithms. A simulated enterprise-level Local Area Network (LAN) was modeled using NetworkX to represent switch-router interconnectivity with redundant links. Fault scenarios, including link failure and congestion, were injected using Scapy, while automatic failover and rerouting were implemented via custom Python logic. The system demonstrates resilience by dynamically detecting path failures, redistributing network traffic through redundant links, and minimizing downtime. Performance evaluations reveal significant improvements in packet delivery continuity, faster recovery times, and reduced packet loss compared to non-fault-tolerant baselines. The implementation provides a scalable and lightweight approach to integrating fault-tolerance features into mid-scale networks, with potential application in enterprise information technology infrastructures and academic simulations.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 22:59:28 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Gbaden",
"Terlumun",
""
],
[
"Ukor",
"Mterorga",
""
],
[
"Ateata",
"Grace Erdoo",
""
]
] |
2508.14306
|
Yuji Ikeda
|
Yuji Ikeda, Axel Forslund, Pranav Kumar, Yongliang Ou, Jong Hyun Jung, Andreas K\"ohn, Blazej Grabowski
|
Machine-learning interatomic potentials achieving CCSD(T) accuracy for van-der-Waals-dominated systems via {\Delta}-learning
| null | null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Machine-learning interatomic potentials (MLIPs) enable large-scale atomistic simulations at moderate computational cost while retaining ab initio accuracy. However, most MLIPs are trained on density-functional theory (DFT), which often falls short of chemical accuracy (1 kcal/mol). Conversely, coupled-cluster methods, particularly CCSD(T), which includes single, double, and perturbative triple excitations, are considered the gold standard of computational chemistry but rarely applied to periodic systems due to their huge computational cost. Here we present a $\Delta$-learning workflow to produce interatomic potentials with CCSD(T) accuracy for periodic systems including van der Waals (vdW) interactions. The procedure combines a dispersion-corrected tight-binding baseline with an MLIP trained on the differences of the target CCSD(T) energies from the baseline. This $\Delta$-learning strategy enables training on compact molecular fragments while preserving transferability. The dispersion interactions are captured by including vdW-bound multimers in the training set; together with the vdW-aware tight-binding baseline, the formally local MLIP attains CCSD(T) accuracy for systems governed by long-range vdW forces. The resulting potential yields root-mean-square energy errors below 0.4 meV/atom on both training and test sets and reproduces electronic total atomization energies, bond lengths, harmonic vibrational frequencies, and inter-molecular interaction energies for benchmark molecular systems. We apply the method to a prototypical quasi-two-dimensional covalent organic framework (COF) composed of carbon and hydrogen. The COF structure, inter-layer binding energies, and hydrogen absorption are analyzed at CCSD(T) accuracy. Overall, the developed $\Delta$-learning approach opens a practical route to large-scale atomistic simulations that include vdW interactions with chemical accuracy.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:02:56 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Ikeda",
"Yuji",
""
],
[
"Forslund",
"Axel",
""
],
[
"Kumar",
"Pranav",
""
],
[
"Ou",
"Yongliang",
""
],
[
"Jung",
"Jong Hyun",
""
],
[
"Köhn",
"Andreas",
""
],
[
"Grabowski",
"Blazej",
""
]
] |
2508.14307
|
Demian Inostroza Amestica
|
Demian Inostroza, Mel Mistica, Ekaterina Vylomova, Chris Guest, Kemal Kurniawan
|
A Joint Multitask Model for Morpho-Syntactic Parsing
|
8 pages, SyntaxFest, UniDive 2025 Morpho-Syntactic Parsing shared task
| null | null | null |
cs.CL
|
http://creativecommons.org/licenses/by/4.0/
|
We present a joint multitask model for the UniDive 2025 Morpho-Syntactic Parsing shared task, where systems predict both morphological and syntactic analyses following novel UD annotation scheme. Our system uses a shared XLM-RoBERTa encoder with three specialized decoders for content word identification, dependency parsing, and morphosyntactic feature prediction. Our model achieves the best overall performance on the shared task's leaderboard covering nine typologically diverse languages, with an average MSLAS score of 78.7 percent, LAS of 80.1 percent, and Feats F1 of 90.3 percent. Our ablation studies show that matching the task's gold tokenization and content word identification are crucial to model performance. Error analysis reveals that our model struggles with core grammatical cases (particularly Nom-Acc) and nominal features across languages.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:09:36 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Inostroza",
"Demian",
""
],
[
"Mistica",
"Mel",
""
],
[
"Vylomova",
"Ekaterina",
""
],
[
"Guest",
"Chris",
""
],
[
"Kurniawan",
"Kemal",
""
]
] |
2508.14308
|
Kyle Kremer
|
Kyle Kremer
|
Compact Objects in Globular Clusters
|
A chapter for the Encyclopedia of Astrophysics (edited by I. Mandel, section editor J. Andrews); 13 pages with 5 figures and 1 table
| null | null | null |
astro-ph.HE astro-ph.GA astro-ph.SR gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
It is now widely established that globular clusters host robust populations of white dwarfs, neutron stars, and black holes throughout their lifetimes. Within clusters, dynamical processes enabled by stellar densities thousands to millions of times larger than typical galactic environments facilitate interactions involving these stellar remnants that give rise to an array of astrophysical phenomena. In particular, stellar clusters have emerged as an important formation site for X-ray sources, radio pulsars, and merging black hole binaries similar to those recently detected as gravitational wave sources by the LIGO/Virgo/KAGRA detectors. This article reviews our current understanding of compact objects in globular clusters, discussing current observational evidence, ways these objects influence the dynamical evolution of their hosts, and future prospects.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:10:09 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Kremer",
"Kyle",
""
]
] |
2508.14309
|
Xiaohai Hu
|
Xiaohai Hu, Jason Laks, Guoxiao Guo, Xu Chen
|
Iterative Youla-Kucera Loop Shaping For Precision Motion Control
|
6pages, To appear at MECC 2025, see https://mecc2025.a2c2.org/
| null | null | null |
eess.SY cs.SY
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
This paper presents a numerically robust approach to multi-band disturbance rejection using an iterative Youla-Kucera parameterization technique. The proposed method offers precise control over shaping the frequency response of a feedback loop while maintaining numerical stability through a systematic design process. By implementing an iterative approach, we overcome a critical numerical issue in rejecting vibrations with multiple frequency bands. Meanwhile, our proposed modification of the all-stabilizing Youla-Kucera architecture enables intuitive design while respecting fundamental performance trade-offs and minimizing undesired waterbed amplifications. Numerical validation on a hard disk drive servo system demonstrates significant performance improvements, enabling enhanced positioning precision for increased storage density. The design methodology extends beyond storage systems to various high-precision control applications where multi-band disturbance rejection is critical.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:12:56 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Hu",
"Xiaohai",
""
],
[
"Laks",
"Jason",
""
],
[
"Guo",
"Guoxiao",
""
],
[
"Chen",
"Xu",
""
]
] |
2508.14310
|
Felix Brandt
|
Felix Brandt, Sun\v{c}ica \v{C}ani\'c, Boris Muha
|
Three-dimensional Navier-Stokes-Biot coupling via a moving reticular plate interface: existence of weak solutions
| null | null | null | null |
math.AP
|
http://creativecommons.org/licenses/by-sa/4.0/
|
We prove the existence of finite-energy weak solutions to a regularized three-dimensional fluid-structure interaction (FSI) problem involving an incompressible, viscous, Newtonian fluid and a multilayered poro(visco)elastic structure. The structure consists of a thick layer modeled by the Biot equations and a thin reticular plate with inertia and elastic energy, transparent to fluid flow. The coupling is nonlinear in the sense that it takes place on a moving interface that is not known a priori but is defined by the solution itself, making the problem a moving-boundary problem. This nonlinear free-boundary coupling, combined with the limited regularity of the Biot displacement, renders the classical weak formulation ill-defined at finite energy. To address this, we introduce a minimally invasive regularization based on a suitable extension and convolution of the Biot displacement, chosen so that the regularized problem remains consistent with the original model. We then construct approximate solutions to the regularized problem via a Lie operator-splitting scheme and derive uniform energy bounds. While these bounds ensure weak and weak* convergence, passing to the limit in the nonlinear terms requires refined compactness arguments, including variants of the Aubin-Lions lemma and tools adapted to moving non-Lipschitz interfaces. The result applies in particular to the purely elastic case (without structural damping) as well as the poroviscoelastic case. This work extends the two-dimensional analysis of Kuan-\v{C}ani\'c-Muha 2024 to the fully three-dimensional setting and, to our knowledge, provides the first existence result for a nonlinearly coupled, multilayer 3D Navier-Stokes-Biot FSI system with a permeable interface.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:22:35 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Brandt",
"Felix",
""
],
[
"Čanić",
"Sunčica",
""
],
[
"Muha",
"Boris",
""
]
] |
2508.14311
|
Quan Zhou
|
Quan Zhou, Jakub Marecek, Robert Shorten
|
Learning Time-Varying Convexifications of Multiple Fairness Measures
| null | null | null | null |
cs.LG cs.AI
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
There is an increasing appreciation that one may need to consider multiple measures of fairness, e.g., considering multiple group and individual fairness notions. The relative weights of the fairness regularisers are a priori unknown, may be time varying, and need to be learned on the fly. We consider the learning of time-varying convexifications of multiple fairness measures with limited graph-structured feedback.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:32:26 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zhou",
"Quan",
""
],
[
"Marecek",
"Jakub",
""
],
[
"Shorten",
"Robert",
""
]
] |
2508.14313
|
Can Jin
|
Can Jin, Yang Zhou, Qixin Zhang, Hongwu Peng, Di Zhang, Marco Pavone, Ligong Han, Zhang-Wei Hong, Tong Che, Dimitris N. Metaxas
|
Your Reward Function for RL is Your Best PRM for Search: Unifying RL and Search-Based TTS
| null | null | null | null |
cs.LG cs.AI
|
http://creativecommons.org/licenses/by/4.0/
|
Test-time scaling (TTS) for large language models (LLMs) has thus far fallen into two largely separate paradigms: (1) reinforcement learning (RL) methods that optimize sparse outcome-based rewards, yet suffer from instability and low sample efficiency; and (2) search-based techniques guided by independently trained, static process reward models (PRMs), which require expensive human- or LLM-generated labels and often degrade under distribution shifts. In this paper, we introduce AIRL-S, the first natural unification of RL-based and search-based TTS. Central to AIRL-S is the insight that the reward function learned during RL training inherently represents the ideal PRM for guiding downstream search. Specifically, we leverage adversarial inverse reinforcement learning (AIRL) combined with group relative policy optimization (GRPO) to learn a dense, dynamic PRM directly from correct reasoning traces, entirely eliminating the need for labeled intermediate process data. At inference, the resulting PRM simultaneously serves as the critic for RL rollouts and as a heuristic to effectively guide search procedures, facilitating robust reasoning chain extension, mitigating reward hacking, and enhancing cross-task generalization. Experimental results across eight benchmarks, including mathematics, scientific reasoning, and code generation, demonstrate that our unified approach improves performance by 9 % on average over the base model, matching GPT-4o. Furthermore, when integrated into multiple search algorithms, our PRM consistently outperforms all baseline PRMs trained with labeled data. These results underscore that, indeed, your reward function for RL is your best PRM for search, providing a robust and cost-effective solution to complex reasoning tasks in LLMs.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:41:15 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Jin",
"Can",
""
],
[
"Zhou",
"Yang",
""
],
[
"Zhang",
"Qixin",
""
],
[
"Peng",
"Hongwu",
""
],
[
"Zhang",
"Di",
""
],
[
"Pavone",
"Marco",
""
],
[
"Han",
"Ligong",
""
],
[
"Hong",
"Zhang-Wei",
""
],
[
"Che",
"Tong",
""
],
[
"Metaxas",
"Dimitris N.",
""
]
] |
2508.14314
|
Aman Goel
|
Aman Goel, Daniel Schwartz, Yanjun Qi
|
Zero-knowledge LLM hallucination detection and mitigation through fine-grained cross-model consistency
| null | null | null | null |
cs.CL cs.AI cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Large language models (LLMs) have demonstrated impressive capabilities across diverse tasks, but they remain susceptible to hallucinations--generating content that appears plausible but contains factual inaccuracies. We present Finch-Zk, a black-box framework that leverages FINe-grained Cross-model consistency to detect and mitigate Hallucinations in LLM outputs without requiring external knowledge sources. Finch-Zk introduces two key innovations: 1) a cross-model consistency checking strategy that reveals fine-grained inaccuracies by comparing responses generated by diverse models from semantically-equivalent prompts, and 2) a targeted mitigation technique that applies precise corrections to problematic segments while preserving accurate content. Experiments on the FELM dataset show Finch-Zk improves hallucination detection F1 scores by 6-39\% compared to existing approaches. For mitigation, Finch-Zk achieves 7-8 absolute percentage points improvement in answer accuracy on the GPQA-diamond dataset when applied to state-of-the-art models like Llama 4 Maverick and Claude 4 Sonnet. Extensive evaluation across multiple models demonstrates that Finch-Zk provides a practical, deployment-ready safeguard for enhancing factual reliability in production LLM systems.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:45:34 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Goel",
"Aman",
""
],
[
"Schwartz",
"Daniel",
""
],
[
"Qi",
"Yanjun",
""
]
] |
2508.14315
|
Pritthijit Nath
|
Pritthijit Nath, Sebastian Schemm, Henry Moss, Peter Haynes, Emily Shuckburgh, Mark Webb
|
FedRAIN-Lite: Federated Reinforcement Algorithms for Improving Idealised Numerical Weather and Climate Models
|
21 pages, 6 figures
| null | null | null |
cs.LG physics.ao-ph
|
http://creativecommons.org/licenses/by-nc-sa/4.0/
|
Sub-grid parameterisations in climate models are traditionally static and tuned offline, limiting adaptability to evolving states. This work introduces FedRAIN-Lite, a federated reinforcement learning (FedRL) framework that mirrors the spatial decomposition used in general circulation models (GCMs) by assigning agents to latitude bands, enabling local parameter learning with periodic global aggregation. Using a hierarchy of simplified energy-balance climate models, from a single-agent baseline (ebm-v1) to multi-agent ensemble (ebm-v2) and GCM-like (ebm-v3) setups, we benchmark three RL algorithms under different FedRL configurations. Results show that Deep Deterministic Policy Gradient (DDPG) consistently outperforms both static and single-agent baselines, with faster convergence and lower area-weighted RMSE in tropical and mid-latitude zones across both ebm-v2 and ebm-v3 setups. DDPG's ability to transfer across hyperparameters and low computational cost make it well-suited for geographically adaptive parameter learning. This capability offers a scalable pathway towards high-complexity GCMs and provides a prototype for physically aligned, online-learning climate models that can evolve with a changing climate. Code accessible at https://github.com/p3jitnath/climate-rl-fedrl.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:54:13 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Nath",
"Pritthijit",
""
],
[
"Schemm",
"Sebastian",
""
],
[
"Moss",
"Henry",
""
],
[
"Haynes",
"Peter",
""
],
[
"Shuckburgh",
"Emily",
""
],
[
"Webb",
"Mark",
""
]
] |
2508.14316
|
Nathan Steinle
|
Nathan Steinle, Samar Safi-Harb
|
Machine learning classification of black holes in the mass--spin diagram
| null | null | null | null |
astro-ph.HE astro-ph.CO gr-qc
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
We present the mass--spin diagram for classifying black holes and studying their formation pathways, providing an analogue to the Hertzsprung-Russell diagram. This allows for black hole evolutionary tracks as a function of redshift, combining formation, accretion, and merger histories for the variety of black hole populations. A realistic black hole continuum constructed from initial mass and spin functions and approximate redshift evolution reveals possible black hole main sequences, such as sustained coherent accretion through cosmic time (i.e., Cosmic Accretion) or hierarchical merger trees. In the stellar-mass regime, we use a binary population synthesis software to compare three spin prescriptions for tidal evolution of Wolf-Rayet progenitors, showing how the mass--spin diagram exposes interesting modeling differences. We then classify black hole populations by applying supervised and unsupervised machine learning clustering methods to mass--spin datasets. While bare unsupervised clustering can nearly recover canonical population boundaries (stellar-mass, intermediate-mass, and supermassive), a more sophisticated approach utilizing deep learning via variational autoencoders for latent space representation learning aids in clustering of realistic datasets with subclasses that highly overlap in mass--spin space. We find that a supervised random forest can accurately recover the correct clusters from the learned latent space representation depending on the complexity of the underlying dataset, semi-supervised methods show potential for further development, and the performance of unsupervised classifiers is a great challenge. Our findings motivate future machine learning applications and demonstrate that the mass--spin diagram can be used to connect gravitational-wave and electromagnetic observations with theoretical models.
|
[
{
"version": "v1",
"created": "Tue, 19 Aug 2025 23:59:05 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Steinle",
"Nathan",
""
],
[
"Safi-Harb",
"Samar",
""
]
] |
2508.14317
|
Zhiheng Yang
|
Jing Chen, Zhiheng Yang, Yixian Shen, Jie Liu, Adam Belloum, Chrysa Papagainni, Paola Grosso
|
SurveyGen-I: Consistent Scientific Survey Generation with Evolving Plans and Memory-Guided Writing
|
The code is available at https://github.com/SurveyGens/SurveyGen-I , 20 pages, 16 figures
| null | null | null |
cs.CL cs.IR
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Survey papers play a critical role in scientific communication by consolidating progress across a field. Recent advances in Large Language Models (LLMs) offer a promising solution by automating key steps in the survey-generation pipeline, such as retrieval, structuring, and summarization. However, existing LLM-based approaches often struggle with maintaining coherence across long, multi-section surveys and providing comprehensive citation coverage. To address these limitations, we introduce SurveyGen-I, an automatic survey generation framework that combines coarse-to-fine retrieval, adaptive planning, and memory-guided generation. SurveyGen-I first performs survey-level retrieval to construct the initial outline and writing plan, and then dynamically refines both during generation through a memory mechanism that stores previously written content and terminology, ensuring coherence across subsections. When the system detects insufficient context, it triggers fine-grained subsection-level retrieval. During generation, SurveyGen-I leverages this memory mechanism to maintain coherence across subsections. Experiments across four scientific domains demonstrate that SurveyGen-I consistently outperforms previous works in content quality, consistency, and citation coverage.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:03:46 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Chen",
"Jing",
""
],
[
"Yang",
"Zhiheng",
""
],
[
"Shen",
"Yixian",
""
],
[
"Liu",
"Jie",
""
],
[
"Belloum",
"Adam",
""
],
[
"Papagainni",
"Chrysa",
""
],
[
"Grosso",
"Paola",
""
]
] |
2508.14319
|
Parshan Javanrood
|
Parshan Javanrood, Matei Ripeanu
|
SSSP-Del: Fully Dynamic Distributed Algorithm for Single-Source Shortest Path
|
Submitted to the IA^3 Workshop at SC25. Under review
| null | null | null |
cs.DC
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Modern graphs are both large and dynamic, presenting significant challenges for fundamental queries, such as the Single-Source Shortest Path (SSSP) problem. Naively recomputing the SSSP tree after each topology change is prohibitively expensive, causing on-demand computation to suffer from high latency. Existing dynamic SSSP algorithms often cannot simultaneously handle both edge additions and deletions, operate in distributed memory, and provide low-latency query results. To address these challenges, this paper presents SSSP-Del, a new vertex-centric, asynchronous, and fully distributed algorithm for dynamic SSSP. Operating in a shared-nothing architecture, our algorithm processes streams of both edge insertions and deletions. We conduct a comprehensive evaluation on large real-world and synthetic graphs with millions of vertices, and provide a thorough analysis by evaluating result latency, solution stability, and throughput.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:07:30 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Javanrood",
"Parshan",
""
],
[
"Ripeanu",
"Matei",
""
]
] |
2508.14320
|
Richard Garner
|
Richard Garner and Jean-Simon Pacaud Lemay
|
Free differential modalities
|
60 pages
| null | null | null |
math.CT
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Categorical models of the exponential modality of linear logic will often, but not always, support an operation of differentiation. When they do, we speak of a monoidal differential modality; when they do not, we have merely a monoidal coalgebra modality. More generally, there are notions of differential modality and coalgebra modality which stand in the same relation, but which move outside the realm of linear logic; they model important structures such as smooth differentiation on Euclidean space.
In this paper, we show that, in a suitably well-behaved k-linear symmetric monoidal category, each (monoidal) coalgebra modality can be freely completed to a (monoidal) differential modality. In particular, we prove the existence of an initial monoidal differential modality, which, even in simple examples such as the category of sets and relations, yields new models of differential linear logic.
Key to our proofs is the concept of an algebraically-free commutative monoid in a symmetric monoidal category. A commutative monoid $M$ is said to be algebraically-free on an object $X$ if actions by the monoid $M$ correspond to self-commuting actions by the mere object $X$. Along the way, we study the theory of algebraically-free commutative monoids and self-commuting actions, and their interplay with differential modality structure.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:11:54 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Garner",
"Richard",
""
],
[
"Lemay",
"Jean-Simon Pacaud",
""
]
] |
2508.14321
|
Mohammad M. Amirian
|
Mohammad M. Amirian, Andrew J. Irwin
|
piCurve: an R package for modeling photosynthesis-irradiance curves
| null | null | null | null |
stat.AP q-bio.QM stat.CO
|
http://creativecommons.org/licenses/by/4.0/
|
Photosynthesis-irradiance (PI) curves are foundational for quantifying primary production, parameterizing ecosystem and biogeochemical models, and interpreting physiological acclimation to light. Despite their broad use, researchers lack a unified, reproducible toolkit to fit, compare, and diagnose the many PI formulations that have accumulated over the last century. We introduce piCurve, an R package that standardizes the modeling of PI relationships, with a library of widely used light-limited, light-saturated, and photoinhibited formulations and a consistent statistical framework for estimation and comparison. With the total of 24 PI models, piCurve supports mean squared error (MSE) and maximum likelihood estimation (MLE), provides uncertainty quantification via information matrix (Hessian), and includes automated, data-informed initialization to improve convergence. Utilities classify PI data into light-limited, light-saturated, and photoinhibited regions, while plotting and 'tidy' helpers streamline workflow and reporting. Together, these features enable reproducible analyses and fair model comparisons, including for curves exhibiting a plateau followed by photoinhibition.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:21:12 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Amirian",
"Mohammad M.",
""
],
[
"Irwin",
"Andrew J.",
""
]
] |
2508.14322
|
Andrew Lucas
|
Isabella Zane, Andrew Lucas
|
Quadrupole-conserving dynamics in the non-commutative plane
|
18 pages, 6 figures
| null | null | null |
cond-mat.stat-mech
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Inspired by ``fracton hydrodynamic" universality classes of dynamics with unusual conservation laws, we present a new dynamical universality class that arises out of local area-preserving dynamics in the non-commutative plane. On this symplectic manifold, the area-preserving spatial symmetry group $\mathrm{SL}(2,\mathbb{R})\rtimes \mathbb{R}^2$ is a symmetry group compatible with non-trivial many-body dynamics. The conservation laws associated to this symmetry group correspond to the dipole and quadrupole moments of the particles. We study the unusual dynamics of a crystal lattice subject to such symmetries, and argue that the hydrodynamic description of lattice dynamics breaks down due to relevant nonlinearities. Numerical simulations of classical Hamiltonian dynamical systems with this symmetry are largely consistent with a tree-level effective field theory estimate for the endpoint of this instability.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:33:36 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zane",
"Isabella",
""
],
[
"Lucas",
"Andrew",
""
]
] |
2508.14323
|
Chen Yixin
|
Yixin Chen, Ying Xiong, Shangyu Wu, Yufei Cui, Xue Liu, Nan Guan, Chun Jason Xue
|
Beyond Semantic Similarity: Reducing Unnecessary API Calls via Behavior-Aligned Retriever
| null | null | null | null |
cs.CL
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Tool-augmented large language models (LLMs) leverage external functions to extend their capabilities, but inaccurate function calls can lead to inefficiencies and increased costs.Existing methods address this challenge by fine-tuning LLMs or using demonstration-based prompting, yet they often suffer from high training overhead and fail to account for inconsistent demonstration samples, which misguide the model's invocation behavior. In this paper, we trained a behavior-aligned retriever (BAR), which provides behaviorally consistent demonstrations to help LLMs make more accurate tool-using decisions. To train the BAR, we construct a corpus including different function-calling behaviors, i.e., calling or non-calling.We use the contrastive learning framework to train the BAR with customized positive/negative pairs and a dual-negative contrastive loss, ensuring robust retrieval of behaviorally consistent examples.Experiments demonstrate that our approach significantly reduces erroneous function calls while maintaining high task performance, offering a cost-effective and efficient solution for tool-augmented LLMs.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:35:50 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Chen",
"Yixin",
""
],
[
"Xiong",
"Ying",
""
],
[
"Wu",
"Shangyu",
""
],
[
"Cui",
"Yufei",
""
],
[
"Liu",
"Xue",
""
],
[
"Guan",
"Nan",
""
],
[
"Xue",
"Chun Jason",
""
]
] |
2508.14325
|
Mohamed AbdulHameed
|
Mohamed AbdulHameed, Khadija Mahbuba, Mahmoud Yaseen, Amr Ibrahim, Daniel Moneghan, Benjamin Beeler
|
Modeling of silver transport in cubic SiC: Integrating molecular dynamics, bounds averaging, and uncertainty quantification
| null | null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Silver released from TRISO fuel particles can migrate through the SiC layer and deposit on reactor components, posing radiation hazards and operational challenges. Despite numerous proposed mechanisms, the precise pathway of silver transport through intact 3C-SiC remains unresolved. We present a physics-informed model for estimating the effective diffusivity of silver in polycrystalline 3C-SiC. Molecular dynamics (MD) simulations yield diffusivities for {\Sigma 3} and {\Sigma 9} grain boundaries (GBs), while literature values are used for other GB types and the bulk. These are combined using a bounds-averaging approach accounting for distinct GB transport properties. Bayesian inference of experimental data provides credible intervals for effective Arrhenius parameters and reveals a correlation between activation energy and pre-exponential factor. Although the homogenized model captures GB-mediated transport mechanisms, it overpredicts silver diffusivity relative to experiments. To resolve this, a multiplicative correction based on reversible trapping at nano-pores is introduced. It is derived from first principles and is shown to reproduce observed transport behavior. Sensitivity analysis identified trap desorption energy and {\Sigma 9} GB diffusivity as dominant factors influencing Ag transport. The resulting framework provides a mechanistic description of Ag transport suitable for integration into higher-scale fuel performance models.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:46:19 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"AbdulHameed",
"Mohamed",
""
],
[
"Mahbuba",
"Khadija",
""
],
[
"Yaseen",
"Mahmoud",
""
],
[
"Ibrahim",
"Amr",
""
],
[
"Moneghan",
"Daniel",
""
],
[
"Beeler",
"Benjamin",
""
]
] |
2508.14326
|
Alexandru Chirv\u{a}situ L.
|
Alexandru Chirvasitu
|
Non-representable quantum measures
|
6 pages + references
| null | null | null |
quant-ph math.FA math.QA
|
http://arxiv.org/licenses/nonexclusive-distrib/1.0/
|
Grade-$d$ measures on a $\sigma$-algebra $\mathcal{A}\subseteq 2^X$ over a set $X$ are generalizations of measures satisfying one of a hierarchy of weak additivity-type conditions initially introduced as interference operators in quantum mechanics. Every signed polymeasure $\lambda$ on $(X,\mathcal{A})^d$ produces a grade-$d$ measure as its diagonal $\widetilde{\lambda}(A):=\lambda(A,\cdots,A)$, and we prove that as soon as $d\ge 2$ measures (as opposed to polymeasures) do not suffice: the separate $\sigma$-additivity of a $\lambda$ producing $\mu=\widetilde{\lambda}$ cannot, generally, be amplified to global $\sigma$-additivity. This amends a result in the literature, asserting the contrary in case $d=2$.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:47:24 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Chirvasitu",
"Alexandru",
""
]
] |
2508.14327
|
Guile Wu
|
Guile Wu, David Huang, Dongfeng Bai, Bingbing Liu
|
MoVieDrive: Multi-Modal Multi-View Urban Scene Video Generation
|
Technical Report
| null | null | null |
cs.CV
|
http://creativecommons.org/licenses/by/4.0/
|
Video generation has recently shown superiority in urban scene synthesis for autonomous driving. Existing video generation approaches to autonomous driving primarily focus on RGB video generation and lack the ability to support multi-modal video generation. However, multi-modal data, such as depth maps and semantic maps, are crucial for holistic urban scene understanding in autonomous driving. Although it is feasible to use multiple models to generate different modalities, this increases the difficulty of model deployment and does not leverage complementary cues for multi-modal data generation. To address this problem, in this work, we propose a novel multi-modal multi-view video generation approach to autonomous driving. Specifically, we construct a unified diffusion transformer model composed of modal-shared components and modal-specific components. Then, we leverage diverse conditioning inputs to encode controllable scene structure and content cues into the unified diffusion model for multi-modal multi-view video generation. In this way, our approach is capable of generating multi-modal multi-view driving scene videos in a unified framework. Our experiments on the challenging real-world autonomous driving dataset, nuScenes, show that our approach can generate multi-modal multi-view urban scene videos with high fidelity and controllability, surpassing the state-of-the-art methods.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:51:36 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Wu",
"Guile",
""
],
[
"Huang",
"David",
""
],
[
"Bai",
"Dongfeng",
""
],
[
"Liu",
"Bingbing",
""
]
] |
2508.14328
|
Jie Gong
|
Jianhang Zhu, Jie Gong
|
Multi-Source Peak Age of Information Optimization in Mobile Edge Computing Systems
|
16 pages, 10 figures, accepted by IEEE Trans. Networking
| null | null | null |
cs.IT math.IT
|
http://creativecommons.org/licenses/by/4.0/
|
Age of Information (AoI) is emerging as a novel metric for measuring information freshness in real-time monitoring systems. For computation-intensive status data, the information is not revealed until being processed. We consider a status update problem in a multi-source single-server system where the sources are scheduled to generate and transmit status data which are received and processed at the edge server. Generate-at-will sources with both random transmission time and process time are considered, introducing the joint optimization of source scheduling and status sampling on the basis of transmission-computation balancing. We show that a random scheduler is optimal for both non-preemptive and preemptive server settings, and the optimal sampler depends on the scheduling result and its structure remains consistent with the single-source system, i.e., threshold-based sampler for non-preemptive case and transmission-aware deterministic sampler for preemptive case. Then, the problem can be transformed to jointly optimizing the scheduling frequencies and the sampling thresholds/functions, which is non-convex. We proposed an alternation optimization algorithm to solve it. Numerical experiments show that the proposed algorithm can achieve the optimal in a wide range of settings.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:56:44 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Zhu",
"Jianhang",
""
],
[
"Gong",
"Jie",
""
]
] |
2508.14329
|
Mohamed AbdulHameed
|
Mohamed AbdulHameed, Anton J. Schneider, Benjamin Beeler, Michael W.D. Cooper
|
Modeling oxygen-void interactions in uranium nitride
| null | null | null | null |
cond-mat.mtrl-sci
|
http://creativecommons.org/licenses/by/4.0/
|
Oxygen impurities in uranium nitride (UN) are reported to influence its swelling behavior under irradiation, yet the underlying mechanism remains unknown. In this work, we develop a first-principles model that quantifies the interaction of oxygen with voids and fission gas bubbles in UN, leading to a reduction in surface energy that can promote swelling. The analysis reveals that segregation of substitutional oxygen at surface nitrogen sites is the primary driver of surface energy reduction, $|\Delta \sigma|$, while oxygen in surface hollow sites plays a minor and sometimes counteracting role. $|\Delta \sigma|$ is most pronounced for small cavities ($R_v$ = 1--10 nm) at intermediate temperatures that coincide with the onset of breakaway swelling in UN. Larger voids require higher temperatures for oxygen adsorption to significantly lower their surface energy. The temperature dependence of $|\Delta \sigma|$ exhibits three regimes: negligible reduction at low temperatures due to sluggish oxygen diffusion, a maximum at intermediate temperatures where oxygen incorporation is optimal, and a decline at high temperatures due to enhanced bulk solubility. A parametric analysis reveals that $|\Delta \sigma|$ depends strongly on both oxygen concentration and cavity size, but is largely insensitive to porosity. Our results suggest that oxygen-induced surface energy reduction is essential for reconciling the mechanistic swelling model of UN with experimental observations.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 00:58:51 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"AbdulHameed",
"Mohamed",
""
],
[
"Schneider",
"Anton J.",
""
],
[
"Beeler",
"Benjamin",
""
],
[
"Cooper",
"Michael W. D.",
""
]
] |
2508.14330
|
N\'icolas Roque Dos Santos
|
N\'icolas Roque dos Santos, Dawon Ahn, Diego Minatel, Alneu de Andrade Lopes, Evangelos E. Papalexakis
|
Multi-view Graph Condensation via Tensor Decomposition
| null | null | null | null |
cs.LG
|
http://creativecommons.org/licenses/by/4.0/
|
Graph Neural Networks (GNNs) have demonstrated remarkable results in various real-world applications, including drug discovery, object detection, social media analysis, recommender systems, and text classification. In contrast to their vast potential, training them on large-scale graphs presents significant computational challenges due to the resources required for their storage and processing. Graph Condensation has emerged as a promising solution to reduce these demands by learning a synthetic compact graph that preserves the essential information of the original one while maintaining the GNN's predictive performance. Despite their efficacy, current graph condensation approaches frequently rely on a computationally intensive bi-level optimization. Moreover, they fail to maintain a mapping between synthetic and original nodes, limiting the interpretability of the model's decisions. In this sense, a wide range of decomposition techniques have been applied to learn linear or multi-linear functions from graph data, offering a more transparent and less resource-intensive alternative. However, their applicability to graph condensation remains unexplored. This paper addresses this gap and proposes a novel method called Multi-view Graph Condensation via Tensor Decomposition (GCTD) to investigate to what extent such techniques can synthesize an informative smaller graph and achieve comparable downstream task performance. Extensive experiments on six real-world datasets demonstrate that GCTD effectively reduces graph size while preserving GNN performance, achieving up to a 4.0\ improvement in accuracy on three out of six datasets and competitive performance on large graphs compared to existing approaches. Our code is available at https://anonymous.4open.science/r/gctd-345A.
|
[
{
"version": "v1",
"created": "Wed, 20 Aug 2025 01:02:18 GMT"
}
] | 2025-08-21T00:00:00 |
[
[
"Santos",
"Nícolas Roque dos",
""
],
[
"Ahn",
"Dawon",
""
],
[
"Minatel",
"Diego",
""
],
[
"Lopes",
"Alneu de Andrade",
""
],
[
"Papalexakis",
"Evangelos E.",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.