id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2507.02580 | Anna Klimova | Anna Klimova, Tam\'as Rudas | On the analysis of sequential designs without a specified number of observations | null | null | null | null | stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The paper focuses on sequential experiments for categorical responses in which whether or not a further observation is made depends on the outcome of a previous experiment. Examples include subsequent medical interventions being performed or not depending on the result of a previous intervention, data about offsprings, life tables, and repeated educational retraininig until a certain proficiency level is achieved. Such experiments do not lead to data with a full Cartesian product structure and, despite a prespecified initial sample size, the total number of observations, or interventions, made cannot be determined in advance. The paper investigates the distributional assumptions behind such data and describes a parameterization of the distribution that arises and the respective model class to analyze it. Both the data structure resulting from such an experiment and the model class are special examples of staged trees in algebraic statistics. The properties of the resulting parameter estimates and test statistics are obtained and illustrated using hypothetical and real data. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:39:01 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Klimova",
"Anna",
""
],
[
"Rudas",
"Tamás",
""
]
] |
2507.02581 | Tan Pan | Tan Pan, Zhaorui Tan, Kaiyu Guo, Dongli Xu, Weidi Xu, Chen Jiang, Xin Guo, Yuan Qi, Yuan Cheng | Structure-aware Semantic Discrepancy and Consistency for 3D Medical Image Self-supervised Learning | Accepted by ICCV25 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | 3D medical image self-supervised learning (mSSL) holds great promise for medical analysis. Effectively supporting broader applications requires considering anatomical structure variations in location, scale, and morphology, which are crucial for capturing meaningful distinctions. However, previous mSSL methods partition images with fixed-size patches, often ignoring the structure variations. In this work, we introduce a novel perspective on 3D medical images with the goal of learning structure-aware representations. We assume that patches within the same structure share the same semantics (semantic consistency) while those from different structures exhibit distinct semantics (semantic discrepancy). Based on this assumption, we propose an mSSL framework named $S^2DC$, achieving Structure-aware Semantic Discrepancy and Consistency in two steps. First, $S^2DC$ enforces distinct representations for different patches to increase semantic discrepancy by leveraging an optimal transport strategy. Second, $S^2DC$ advances semantic consistency at the structural level based on neighborhood similarity distribution. By bridging patch-level and structure-level representations, $S^2DC$ achieves structure-aware representations. Thoroughly evaluated across 10 datasets, 4 tasks, and 3 modalities, our proposed method consistently outperforms the state-of-the-art methods in mSSL. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:40:46 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Pan",
"Tan",
""
],
[
"Tan",
"Zhaorui",
""
],
[
"Guo",
"Kaiyu",
""
],
[
"Xu",
"Dongli",
""
],
[
"Xu",
"Weidi",
""
],
[
"Jiang",
"Chen",
""
],
[
"Guo",
"Xin",
""
],
[
"Qi",
"Yuan",
""
],
[
"Cheng",
"Yuan",
""
]
] |
2507.02582 | Pavel Naumov | Junli Jiang and Pavel Naumov | Responsibility Gap and Diffusion in Sequential Decision-Making Mechanisms | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Responsibility has long been a subject of study in law and philosophy. More recently, it became a focus of AI literature. The article investigates the computational complexity of two important properties of responsibility in collective decision-making: diffusion and gap. It shows that the sets of diffusion-free and gap-free decision-making mechanisms are $\Pi_2$-complete and $\Pi_3$-complete, respectively. At the same time, the intersection of these classes is $\Pi_2$-complete. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:43:38 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Jiang",
"Junli",
""
],
[
"Naumov",
"Pavel",
""
]
] |
2507.02583 | Swarnim Shashank | Swarnim Shashank, Askar B. Abdikamalov, Honghui Liu, Abdurakhmon Nosirov, Cosimo Bambi, Indu K. Dihingia, Yosuke Mizuno | Measuring black hole spins with X-ray reflection spectroscopy: A GRMHD outlook | 18 pages, 10 figures; supplemental material available upon request | null | null | null | astro-ph.HE gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | X-ray reflection spectroscopy has evolved as one of the leading methods to measure black hole spins. However, the question is whether its measurements are subjected to systematic biases, especially considering the possible discrepancy between the spin measurements inferred with this technique and those from gravitational wave observations. In this work, we use general relativistic magnetohydrodynamic (GRMHD) simulations of thin accretion disks around spinning black holes for modeling the accretion process, and then we simulate NuSTAR observations to test the capability of modern reflection models in recovering the input spins. For the first time, we model the electron density and ionization profiles from GRMHD-simulated disks. Our study reveals that current reflection models work well only for fast-rotating black holes. We model the corona as the base of the jet and we find that reflection models with lamppost emissivity profiles fail to recover the correct black hole spins. Reflection models with broken power-law emissivity profiles perform better. As we increase the complexity of the simulated models, it is more difficult to recover the correct input spins, pointing towards the need to update our current reflection models with more advanced accretion disks and coronal geometries. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:44:57 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shashank",
"Swarnim",
""
],
[
"Abdikamalov",
"Askar B.",
""
],
[
"Liu",
"Honghui",
""
],
[
"Nosirov",
"Abdurakhmon",
""
],
[
"Bambi",
"Cosimo",
""
],
[
"Dihingia",
"Indu K.",
""
],
[
"Mizuno",
"Yosuke",
""
]
] |
2507.02584 | Wenwei Que | Wenwei Que and Yang Li and Lu Wang and Wentao Liu and Yougang Bian and Manjiang Hu and Yongfu Li | Observer-Based Distributed Model Predictive Control for String-Stable Multi-vehicle Systems with Markovian Switching Topology | 8 pages,7 figures,conference | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Switching communication topologies can cause instability in vehicle platoons, as vehicle information may be lost during the dynamic switching process. This highlights the need to design a controller capable of maintaining the stability of vehicle platoons under dynamically changing topologies. However, capturing the dynamic characteristics of switching topologies and obtaining complete vehicle information for controller design while ensuring stability remains a significant challenge. In this study, we propose an observer-based distributed model predictive control (DMPC) method for vehicle platoons under directed Markovian switching topologies. Considering the stochastic nature of the switching topologies, we model the directed switching communication topologies using a continuous-time Markov chain. To obtain the leader vehicle's information for controller design, we develop a fully distributed adaptive observer that can quickly adapt to the randomly switching topologies, ensuring that the observed information is not affected by the dynamic topology switches. Additionally, a sufficient condition is derived to guarantee the mean-square stability of the observer. Furthermore, we construct the DMPC terminal update law based on the observer and formulate a string stability constraint based on the observed information. Numerical simulations demonstrate that our method can reduce tracking errors while ensuring string stability. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:45:21 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Que",
"Wenwei",
""
],
[
"Li",
"Yang",
""
],
[
"Wang",
"Lu",
""
],
[
"Liu",
"Wentao",
""
],
[
"Bian",
"Yougang",
""
],
[
"Hu",
"Manjiang",
""
],
[
"Li",
"Yongfu",
""
]
] |
2507.02585 | Fabian Kresse | Fabian Kresse and Emily Yu and Christoph H. Lampert | Scalable Interconnect Learning in Boolean Networks | 12 pages, 8 Figures | null | null | null | cs.LG cs.LO | http://creativecommons.org/licenses/by-sa/4.0/ | Learned Differentiable Boolean Logic Networks (DBNs) already deliver efficient inference on resource-constrained hardware. We extend them with a trainable, differentiable interconnect whose parameter count remains constant as input width grows, allowing DBNs to scale to far wider layers than earlier learnable-interconnect designs while preserving their advantageous accuracy. To further reduce model size, we propose two complementary pruning stages: an SAT-based logic equivalence pass that removes redundant gates without affecting performance, and a similarity-based, data-driven pass that outperforms a magnitude-style greedy baseline and offers a superior compression-accuracy trade-off. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:45:45 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kresse",
"Fabian",
""
],
[
"Yu",
"Emily",
""
],
[
"Lampert",
"Christoph H.",
""
]
] |
2507.02586 | Francesco D'Amico | Flavio Nicoletti, Francesco D'Amico, Matteo Negri | Statistical mechanics of vector Hopfield network near and above saturation | 38 pages, 10 figures | null | null | null | cond-mat.dis-nn | http://creativecommons.org/licenses/by/4.0/ | We study analytically and numerically a Hopfield fully-connected network with $d$-vector spins. These networks are models of associative memory that generalize the standard Hopfield with Ising spins. We study the equilibrium and out-of-equilibrium properties of the system, considering the system in its retrieval phase $\alpha<\alpha_c$ and beyond. We derive the Replica Symmetric solution for the equilibrium thermodynamics of the system, together with its phase diagram: we find that the retrieval phase of the network shrinks with growing spin dimension, having ultimately a vanishing critical capacity $\alpha_c\propto 1/d$ in the large $d$ limit. As a trade-off, we observe that in the same limit vector Hopfield networks are able to denoise corrupted input patterns in the first step of retrieval dynamics, up to very large capacities $\widetilde{\alpha}\propto d$. We also study the static properties of the system at zero temperature, considering the statistical properties of soft modes of the energy Hessian spectrum. We find that local minima of the energy landscape related to memory states have ungapped spectra with rare soft eigenmodes: these excitations are localized, their measure condensating on the noisiest neurons of the memory state. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:48:34 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Nicoletti",
"Flavio",
""
],
[
"D'Amico",
"Francesco",
""
],
[
"Negri",
"Matteo",
""
]
] |
2507.02587 | Garnik F. Mkrtchian | H. K. Avetissian and G. F. Mkrtchian | Zeptosecond Electron Pulse Train and Ultrafast Coherent Control of Quantum States via Multiphoton Inelastic Cherenkov Diffraction | 10 pages, 8 figures | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | We investigate the quantum dynamics of fermionic particles interacting with a laser field in a gaseous medium, in the regime of inelastic diffraction scattering on the phase lattice of a slowed travelling wave, below the critical field of induced Cherenkov process. Using a relativistic quantum kinetic framework and numerical solutions of Dirac equation in the rest frame of the slowed wave, we analyze the evolution of actual electron wave packets and beams at the inelastic scattering on the actual laser pulses of finite duration. Our results reveal coherent multiphoton exchange involving up to $10^{4}$ photons and the emergence of attosecond-zeptosecond electron sub-bunches after the free-space propagation. The pulse compression by such mechanism is robust to laser pulse duration but sensitive to the initial momentum spread of the particles/beams. We propose a mechanism to achieve electron pulses in zeptosecond time scales with potentiality for ultrafast coherent control of quantum states that opens new avenues in high-resolution temporal structuring of electron beams for time-resolved quantum technologies and attosecond-zeptosecond science, as well as, for application in high-resolution electron microscopy. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:49:11 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Avetissian",
"H. K.",
""
],
[
"Mkrtchian",
"G. F.",
""
]
] |
2507.02588 | Elsa Maneval | Elsa Maneval | Non-archimedean topological mirror symmetry for $SL_n$ and $PGL_n$ Higgs bundles | 34 pages, comments welcome | null | null | null | math.AG | http://creativecommons.org/licenses/by/4.0/ | The Hausel-Thaddeus conjectures concern topological mirror symmetry between moduli spaces of $SL_n$ and $PGL_n$ Higgs bundles on a curve. A non-archimedean approach was introduced by Groechenig, Wyss and Ziegler, proving the conjecture for coprime rank and degree. This article is concerned with its generalisation to the non-coprime case. We treat both the classical ($D=K$) and meromorphic ($D>K$) settings. We prove an equality of $p$-adic volumes twisted by gerbes between moduli spaces of $SL_n$ and $PGL_n$ Higgs bundles of arbitrary degree. In the meromorphic case, building on results of Maulik and Shen, we show that these twisted $p$-adic volumes are related to intersection cohomology. We also conjecture a connection between these $p$-adic volumes and $BPS$ cohomology. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:49:39 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Maneval",
"Elsa",
""
]
] |
2507.02589 | Giulio Belletti | Giulio Belletti, Renaud Detcherry | An effective proof of finiteness for Kauffman bracket skein modules | 44 pages, 10 figures. Comments welcome | null | null | null | math.GT | http://creativecommons.org/licenses/by/4.0/ | We prove a version of the finiteness conjecture for Kauffman bracket skein modules of $3$-manifolds with boundary, which was introduced by the second author in \cite{Det21}. In particular our methods, which are constructive, give an alternative proof of Witten's finiteness conjecture for the Kauffman bracket skein modules of closed $3$-manifolds, which was originally proved in \cite{GJS19}. Moreover, as a corollary we show that the peripheral ideal of any link is non-empty, answering a question of Frohman, Gelca and Lofaro \cite{FGL02}. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:51:51 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Belletti",
"Giulio",
""
],
[
"Detcherry",
"Renaud",
""
]
] |
2507.02590 | Youssef Rami | M. T. K. Abassi, A. Acharqy, K. Boutahir, and Y. Rami | An algorithm to determine LS-category and Ginsburg invariant of any rationally elliptic space | 36 pages | null | null | null | math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $X$ be a rationally elliptic space. Utilizing the Gorenstein algebra structure of $X$, we present three algorithms that together induce a generating class of $Ext^N_{(\Lambda V,d)}(\mathbb{Q},(\Lambda V,d))$ with $N$ being the formal dimension of $X$. From these algorithms, we derive an algorithm to compute the rational Lusternik-Schnirelmann category $cat_0(X)$. Furthermore, by applying a spectral sequence argument based on the {\it Eilenberg-Moore spectral sequence}, we compute the rational Ginsburg invariant $l_0(X)$ introduced by M. Ginsburg in \cite{Gin}. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:53:25 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Abassi",
"M. T. K.",
""
],
[
"Acharqy",
"A.",
""
],
[
"Boutahir",
"K.",
""
],
[
"Rami",
"Y.",
""
]
] |
2507.02591 | Weili Xu | Weili Xu, Enxin Song, Wenhao Chai, Xuexiang Wen, Tian Ye, Gaoang Wang | AuroraLong: Bringing RNNs Back to Efficient Open-Ended Video Understanding | Accepted to ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The challenge of long video understanding lies in its high computational complexity and prohibitive memory cost, since the memory and computation required by transformer-based LLMs scale quadratically with input sequence length. We propose AuroraLong to address this challenge by replacing the LLM component in MLLMs with a linear RNN language model that handles input sequence of arbitrary length with constant-size hidden states. To further increase throughput and efficiency, we combine visual token merge with linear RNN models by reordering the visual tokens by their sizes in ascending order. Despite having only 2B parameters and being trained exclusively on public data, AuroraLong achieves performance comparable to Transformer-based models of similar size trained on private datasets across multiple video benchmarks. This demonstrates the potential of efficient, linear RNNs to democratize long video understanding by lowering its computational entry barrier. To our best knowledge, we are the first to use a linear RNN based LLM backbone in a LLaVA-like model for open-ended video understanding. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:55:16 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Xu",
"Weili",
""
],
[
"Song",
"Enxin",
""
],
[
"Chai",
"Wenhao",
""
],
[
"Wen",
"Xuexiang",
""
],
[
"Ye",
"Tian",
""
],
[
"Wang",
"Gaoang",
""
]
] |
2507.02592 | Kuan Li | Kuan Li, Zhongwang Zhang, Huifeng Yin, Liwen Zhang, Litu Ou, Jialong Wu, Wenbiao Yin, Baixuan Li, Zhengwei Tao, Xinyu Wang, Weizhou Shen, Junkai Zhang, Dingchu Zhang, Xixi Wu, Yong Jiang, Ming Yan, Pengjun Xie, Fei Huang, Jingren Zhou | WebSailor: Navigating Super-human Reasoning for Web Agent | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/publicdomain/zero/1.0/ | Transcending human cognitive limitations represents a critical frontier in LLM training. Proprietary agentic systems like DeepResearch have demonstrated superhuman capabilities on extremely complex information-seeking benchmarks such as BrowseComp, a feat previously unattainable. We posit that their success hinges on a sophisticated reasoning pattern absent in open-source models: the ability to systematically reduce extreme uncertainty when navigating vast information landscapes. Based on this insight, we introduce WebSailor, a complete post-training methodology designed to instill this crucial capability. Our approach involves generating novel, high-uncertainty tasks through structured sampling and information obfuscation, RFT cold start, and an efficient agentic RL training algorithm, Duplicating Sampling Policy Optimization (DUPO). With this integrated pipeline, WebSailor significantly outperforms all opensource agents in complex information-seeking tasks, matching proprietary agents' performance and closing the capability gap. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:59:07 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Li",
"Kuan",
""
],
[
"Zhang",
"Zhongwang",
""
],
[
"Yin",
"Huifeng",
""
],
[
"Zhang",
"Liwen",
""
],
[
"Ou",
"Litu",
""
],
[
"Wu",
"Jialong",
""
],
[
"Yin",
"Wenbiao",
""
],
[
"Li",
"Baixuan",
""
],
[
"Tao",
"Zhengwei",
""
],
[
"Wang",
"Xinyu",
""
],
[
"Shen",
"Weizhou",
""
],
[
"Zhang",
"Junkai",
""
],
[
"Zhang",
"Dingchu",
""
],
[
"Wu",
"Xixi",
""
],
[
"Jiang",
"Yong",
""
],
[
"Yan",
"Ming",
""
],
[
"Xie",
"Pengjun",
""
],
[
"Huang",
"Fei",
""
],
[
"Zhou",
"Jingren",
""
]
] |
2507.02593 | Matthias A{\ss}enmacher | Cornelia Gruber and Helen Alber and Bernd Bischl and G\"oran Kauermann and Barbara Plank and Matthias A{\ss}enmacher | Revisiting Active Learning under (Human) Label Variation | null | null | null | null | cs.CL cs.HC cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Access to high-quality labeled data remains a limiting factor in applied supervised learning. While label variation (LV), i.e., differing labels for the same instance, is common, especially in natural language processing, annotation frameworks often still rest on the assumption of a single ground truth. This overlooks human label variation (HLV), the occurrence of plausible differences in annotations, as an informative signal. Similarly, active learning (AL), a popular approach to optimizing the use of limited annotation budgets in training ML models, often relies on at least one of several simplifying assumptions, which rarely hold in practice when acknowledging HLV. In this paper, we examine foundational assumptions about truth and label nature, highlighting the need to decompose observed LV into signal (e.g., HLV) and noise (e.g., annotation error). We survey how the AL and (H)LV communities have addressed -- or neglected -- these distinctions and propose a conceptual framework for incorporating HLV throughout the AL loop, including instance selection, annotator choice, and label representation. We further discuss the integration of large language models (LLM) as annotators. Our work aims to lay a conceptual foundation for HLV-aware active learning, better reflecting the complexities of real-world annotation. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 12:59:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Gruber",
"Cornelia",
""
],
[
"Alber",
"Helen",
""
],
[
"Bischl",
"Bernd",
""
],
[
"Kauermann",
"Göran",
""
],
[
"Plank",
"Barbara",
""
],
[
"Aßenmacher",
"Matthias",
""
]
] |
2507.02594 | Morteza Baniasad Azad | Morteza Baniasad Azad and Mostafa Arabtash | Recognition by the set of exponents in the prime factorization of the product of element orders | null | null | null | null | math.GR | http://creativecommons.org/licenses/by/4.0/ | Let $G$ be a finite group. Let $\rho(G) = \prod_{g \in G} o(g)={p_1}^{\alpha_1} {p_2}^{\alpha_2} \cdots {p_k}^{\alpha_k}$, where $p_1, p_2, \cdots, p_k$ are distinct prime numbers and $o(g)$ denotes the order of $g \in G$. The set of exponents in the prime factorization of the product of element orders is denoted by $ {\operatorname{Exp}}_{\rho}(G)$, i.e., $ {\operatorname{Exp}}_{\rho}(G)=\{\alpha_1,\alpha_2, \cdots,\alpha_k\}$. In this paper, we give a new characterization for some groups by $ {\operatorname{Exp}}_{\rho}(G)$. We prove that the groups ${\rm PSL}(2, 5) \times \mathbb{Z}_p$, ${\rm PSL}(2, 7)$ and ${\rm PSL}(2, 11)$ are uniquely determined by $ {\operatorname{Exp}}_{\rho}(G)$. Furthermore, we prove that the groups ${\rm PSL}(2, 5)$ and ${\rm PSL}(2, 13)$ are uniquely determined by the parameters $ {\operatorname{Exp}}_{\rho}(G)$ and $|G|$. Additionally, we prove that if ${\operatorname{Exp}}_{\rho}(G) = {\operatorname{Exp}}_{\rho}(\mathbb{Z}_{2qr})$, then $G \cong {\rm PSL}(2, 5)$ or $G \cong \mathbb{Z}_{2qr}$, where $q$ and $r$ are distinct odd prime numbers. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:06:41 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Azad",
"Morteza Baniasad",
""
],
[
"Arabtash",
"Mostafa",
""
]
] |
2507.02595 | Xin Guan | Xin Guan, PeiHsin Lin, Zekun Wu, Ze Wang, Ruibo Zhang, Emre Kazim, Adriano Koshiyama | MPF: Aligning and Debiasing Language Models post Deployment via Multi Perspective Fusion | Accepted at ICML 2025 AIW Workshop | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Multiperspective Fusion (MPF) is a novel posttraining alignment framework for large language models (LLMs) developed in response to the growing need for easy bias mitigation. Built on top of the SAGED pipeline, an automated system for constructing bias benchmarks and extracting interpretable baseline distributions, MPF leverages multiperspective generations to expose and align biases in LLM outputs with nuanced, humanlike baselines. By decomposing baseline, such as sentiment distributions from HR professionals, into interpretable perspective components, MPF guides generation through sampling and balancing of responses, weighted by the probabilities obtained in the decomposition. Empirically, we demonstrate its ability to align LLM sentiment distributions with both counterfactual baselines (absolute equality) and the HR baseline (biased for Top Univeristy), resulting in small KL divergence, reduction of calibration error and generalization to unseen questions. This shows that MPF offers a scalable and interpretable method for alignment and bias mitigation, compatible with deployed LLMs and requiring no extensive prompt engineering or finetuning. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:09:18 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Guan",
"Xin",
""
],
[
"Lin",
"PeiHsin",
""
],
[
"Wu",
"Zekun",
""
],
[
"Wang",
"Ze",
""
],
[
"Zhang",
"Ruibo",
""
],
[
"Kazim",
"Emre",
""
],
[
"Koshiyama",
"Adriano",
""
]
] |
2507.02596 | Tatsuaki Tsuruyama | Tatsuaki Tsuruyama | Relativistic Limits of Decoding: Critical Divergence of Kullback-Leibler Information and Free Energy | null | null | null | null | math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We present a statistical mechanical framework based on the Kullback-Leibler divergence (KLD) to analyze the relativistic limits of decoding time-encoded information from a moving source. By modeling the symbol durations as entropy-maximizing sequences and treating the decoding process as context-sensitive inference, we identify KLD between the sender and receiver distributions as a key indicator of contextual mismatch. We show that, under Lorentz transformations, this divergence grows with the sender's velocity and exhibits critical divergence as the velocity approaches the speed of light. Furthermore, we derive an analytic expression for the Fisher information and demonstrate that decoding sensitivity diverges similarly, indicating instability near the relativistic limit. By introducing an information-theoretic free energy based on the decoding cost, we determine a critical velocity beyond which decoding becomes thermodynamically impossible. These results reveal a phase-transition-like behavior in relativistic information transfer and provide a unified interpretation of KLD, Fisher information, and free energy as measures of decodability. The formalism developed here offers new insights into high-speed communication, relativistic signal processing, and information geometry in non-inertial frames. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:20:00 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Tsuruyama",
"Tatsuaki",
""
]
] |
2507.02597 | J\'er\^ome B\'etrisey | J\'er\^ome B\'etrisey, Anne-Marie Broomhall, Sylvain N. Breton, Rafael A. Garc\'ia, Henry Davenport, Oleg Kochukhov | Frequency separation ratios do not suppress magnetic activity effects in solar-like stars | Accepted for publication as a Letter in Astronomy and Astrophysics | null | null | null | astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Magnetic activity effects are typically neglected in asteroseismic modelling of solar-type stars, presuming that these effects can be accounted for in the parametrisation of the surface effects. It was however demonstrated that magnetic activity can have a significant impact on the asteroseismic characterisation using both forward and inverse techniques. We investigated whether frequency separation ratios, which are commonly used to efficiently suppress surface effects, are also able to suppress magnetic activity effects. Based on GOLF and BiSON observations of the Sun-as-a-star, we performed asteroseismic characterisations using frequency separation ratios as constraints to measure the apparent temporal evolution of the stellar parameters and their correlation with the 10.7 cm radio flux. Frequency separation ratios do not suppress the effects of magnetic activity. Both $r_{01}$ and $r_{02}$ ratios exhibit a clear signature of the magnetic activity cycle. Consequently, when these ratios are employed as constraints in asteroseismic modelling, magnetic activity effects are propagated to the stellar characterisation. Additionally, most stellar parameters correlate with the activity cycle, unlike the direct fitting of individual frequencies. Magnetic activity effects significantly impact asteroseismic characterisation, regardless of whether forward modelling or inverse methods are used. Standard techniques to suppress surface effects have proven ineffective against magnetic activity influences and systematic uncertainties of 4.7%, 2.9%, and 1.0% should be considered for the stellar age, mass, and radius, respectively. In preparation for future space-based photometry missions, it is therefore essential to enhance our theoretical understanding of these effects and develop a modelling procedure capable of accounting for or efficiently suppressing them. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:20:41 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Bétrisey",
"Jérôme",
""
],
[
"Broomhall",
"Anne-Marie",
""
],
[
"Breton",
"Sylvain N.",
""
],
[
"García",
"Rafael A.",
""
],
[
"Davenport",
"Henry",
""
],
[
"Kochukhov",
"Oleg",
""
]
] |
2507.02598 | Chenhao Xue | Chenhao Xue, Kezhi Li, Jiaxing Zhang, Yi Ren, Zhengyuan Shi, Chen Zhang, Yibo Lin, Lining Zhang, Qiang Xu, Guangyu Sun | AC-Refiner: Efficient Arithmetic Circuit Optimization Using Conditional Diffusion Models | 8 pages, 12 figures | null | null | null | cs.AR cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Arithmetic circuits, such as adders and multipliers, are fundamental components of digital systems, directly impacting the performance, power efficiency, and area footprint. However, optimizing these circuits remains challenging due to the vast design space and complex physical constraints. While recent deep learning-based approaches have shown promise, they struggle to consistently explore high-potential design variants, limiting their optimization efficiency. To address this challenge, we propose AC-Refiner, a novel arithmetic circuit optimization framework leveraging conditional diffusion models. Our key insight is to reframe arithmetic circuit synthesis as a conditional image generation task. By carefully conditioning the denoising diffusion process on target quality-of-results (QoRs), AC-Refiner consistently produces high-quality circuit designs. Furthermore, the explored designs are used to fine-tune the diffusion model, which focuses the exploration near the Pareto frontier. Experimental results demonstrate that AC-Refiner generates designs with superior Pareto optimality, outperforming state-of-the-art baselines. The performance gain is further validated by integrating AC-Refiner into practical applications. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:21:33 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Xue",
"Chenhao",
""
],
[
"Li",
"Kezhi",
""
],
[
"Zhang",
"Jiaxing",
""
],
[
"Ren",
"Yi",
""
],
[
"Shi",
"Zhengyuan",
""
],
[
"Zhang",
"Chen",
""
],
[
"Lin",
"Yibo",
""
],
[
"Zhang",
"Lining",
""
],
[
"Xu",
"Qiang",
""
],
[
"Sun",
"Guangyu",
""
]
] |
2507.02599 | Sertac Kilickaya | Sertac Kilickaya and Levent Eren | Pad\'e Approximant Neural Networks for Enhanced Electric Motor Fault Diagnosis Using Vibration and Acoustic Data | Submitted to the Journal of Vibration Engineering & Technologies | null | null | null | cs.LG cs.SD cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Purpose: The primary aim of this study is to enhance fault diagnosis in induction machines by leveraging the Pad\'e Approximant Neuron (PAON) model. While accelerometers and microphones are standard in motor condition monitoring, deep learning models with nonlinear neuron architectures offer promising improvements in diagnostic performance. This research addresses the question: Can Pad\'e Approximant Neural Networks (Pad\'eNets) outperform conventional Convolutional Neural Networks (CNNs) and Self-Organized Operational Neural Networks (Self-ONNs) in diagnosing electrical and mechanical faults using vibration and acoustic data?
Methods: We evaluate and compare the diagnostic capabilities of three deep learning architectures: one-dimensional CNNs, Self-ONNs, and Pad\'eNets. These models are tested on the University of Ottawa's publicly available constant-speed induction motor datasets, which include both vibration and acoustic sensor data. The Pad\'eNet model is designed to introduce enhanced nonlinearity and is compatible with unbounded activation functions such as Leaky ReLU.
Results and Conclusion: Pad\'eNets consistently outperformed the baseline models, achieving diagnostic accuracies of 99.96%, 98.26%, 97.61%, and 98.33% for accelerometers 1, 2, 3, and the acoustic sensor, respectively. The enhanced nonlinearity of Pad\'eNets, together with their compatibility with unbounded activation functions, significantly improves fault diagnosis performance in induction motor condition monitoring. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:21:45 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kilickaya",
"Sertac",
""
],
[
"Eren",
"Levent",
""
]
] |
2507.02600 | Qiaojun Yu | Qiaojun Yu, Xibin Yuan, Yu jiang, Junting Chen, Dongzhe Zheng, Ce Hao, Yang You, Yixing Chen, Yao Mu, Liu Liu, Cewu Lu | ArtGS:3D Gaussian Splatting for Interactive Visual-Physical Modeling and Manipulation of Articulated Objects | Accepted by IROS 2025 | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Articulated object manipulation remains a critical challenge in robotics due to the complex kinematic constraints and the limited physical reasoning of existing methods. In this work, we introduce ArtGS, a novel framework that extends 3D Gaussian Splatting (3DGS) by integrating visual-physical modeling for articulated object understanding and interaction. ArtGS begins with multi-view RGB-D reconstruction, followed by reasoning with a vision-language model (VLM) to extract semantic and structural information, particularly the articulated bones. Through dynamic, differentiable 3DGS-based rendering, ArtGS optimizes the parameters of the articulated bones, ensuring physically consistent motion constraints and enhancing the manipulation policy. By leveraging dynamic Gaussian splatting, cross-embodiment adaptability, and closed-loop optimization, ArtGS establishes a new framework for efficient, scalable, and generalizable articulated object modeling and manipulation. Experiments conducted in both simulation and real-world environments demonstrate that ArtGS significantly outperforms previous methods in joint estimation accuracy and manipulation success rates across a variety of articulated objects. Additional images and videos are available on the project website: https://sites.google.com/view/artgs/home | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:22:18 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yu",
"Qiaojun",
""
],
[
"Yuan",
"Xibin",
""
],
[
"jiang",
"Yu",
""
],
[
"Chen",
"Junting",
""
],
[
"Zheng",
"Dongzhe",
""
],
[
"Hao",
"Ce",
""
],
[
"You",
"Yang",
""
],
[
"Chen",
"Yixing",
""
],
[
"Mu",
"Yao",
""
],
[
"Liu",
"Liu",
""
],
[
"Lu",
"Cewu",
""
]
] |
2507.02601 | Keiji Matsumoto | Keiji Matsumoto | More on Intractability of Thermalization: (almost) i.i.d. inputs and finite lattices | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | This work is an extention of Shiraishi and Matsumoto [10], and discusses the computational complexity of the long-term average of local observables in one-dimensional lattices with shift-invariant nearest-neighbor interactions for simple initial states. As shown in the previous paper, the problem is generally intractable. In this paper we refine the statement further. First, we consider restriction of the initial state, where the state of all the sites are the same except for a single site. We show this version of the problem is also undecidable (RE-complete). Then we turn to the case where the lattice size is finite: depening on the defitiniton of the input size, this version of problem is either EXPSPACE-complete or PSPACE-complete. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:22:33 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Matsumoto",
"Keiji",
""
]
] |
2507.02602 | Riccardo Gallon | Riccardo Gallon, Fabian Schiemenz, Alessandra Menicucci, Eberhard Gill | Addressing Camera Sensors Faults in Vision-Based Navigation: Simulation and Dataset Development | Submitted to Acta Astronautica | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | The increasing importance of Vision-Based Navigation (VBN) algorithms in space missions raises numerous challenges in ensuring their reliability and operational robustness. Sensor faults can lead to inaccurate outputs from navigation algorithms or even complete data processing faults, potentially compromising mission objectives. Artificial Intelligence (AI) offers a powerful solution for detecting such faults, overcoming many of the limitations associated with traditional fault detection methods. However, the primary obstacle to the adoption of AI in this context is the lack of sufficient and representative datasets containing faulty image data.
This study addresses these challenges by focusing on an interplanetary exploration mission scenario. A comprehensive analysis of potential fault cases in camera sensors used within the VBN pipeline is presented. The causes and effects of these faults are systematically characterized, including their impact on image quality and navigation algorithm performance, as well as commonly employed mitigation strategies. To support this analysis, a simulation framework is introduced to recreate faulty conditions in synthetically generated images, enabling a systematic and controlled reproduction of faulty data. The resulting dataset of fault-injected images provides a valuable tool for training and testing AI-based fault detection algorithms. The final link to the dataset will be added after an embargo period. For peer-reviewers, this private link is available. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:23:22 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Gallon",
"Riccardo",
""
],
[
"Schiemenz",
"Fabian",
""
],
[
"Menicucci",
"Alessandra",
""
],
[
"Gill",
"Eberhard",
""
]
] |
2507.02603 | Thomas Lindvall | T. Lindvall, K. J. Hanhij\"arvi, T. Fordell, and A. E. Wallin | Measurement of the Differential Static Scalar Polarizability of the $^\mathbf{88}$Sr$^{+}$ Clock Transition | To appear in Phys. Rev. Lett. 7 pages, 6 figures, plus Supplemental Material 3 pages, 1 figure | null | 10.1103/52by-28mr | null | physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | We report on a precision measurement of the differential static scalar polarizability $\Delta\alpha_0$ of the ${}^{2}\!S_{1/2} \rightarrow {}^{2}\!D_{5/2}$ optical clock transition in the $^{88}$Sr$^{+}$ ion. The polarizability was determined from the 'magic' ion-trap drive frequency where the micromotion-induced second-order Doppler and quadratic Stark shifts cancel, using a single clock in an interleaved scheme by switching between minimized and large micromotion. By measuring at different Mathieu $q_z$ parameters, $\Delta\alpha_0$ can be obtained without prior knowledge of the angle between the rf electric field and the trap axis, which would otherwise dominate the systematic uncertainty. For validation, measurements were carried out at different micromotion levels and by displacing the ion in opposite directions. The results show excellent consistency and our value, $\Delta\alpha_0 = -4.8314(20)\times 10^{-40}\;\mathrm{J\, m^2/V^2} = -29.303(12)\;\mathrm{au}$, reduces the uncertainty by a factor of 3.5 compared to a previous measurement, while showing a discrepancy of $5 \sigma$. Our measurement reduces the polarizability-related uncertainty of the $^{88}$Sr$^{+}$ clock to $2.2\times 10^{-19}$ for a blackbody radiation temperature of 295 K -- a significant step towards total uncertainties ${<}1\times10^{-18}$. Using existing polarizability transfer schemes, the result can reduce the uncertainty also for other ion species. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:25:17 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Lindvall",
"T.",
""
],
[
"Hanhijärvi",
"K. J.",
""
],
[
"Fordell",
"T.",
""
],
[
"Wallin",
"A. E.",
""
]
] |
2507.02604 | Eleanor Hamilton | Eleanor Hamilton, Marta Colleoni, Jonathan E. Thompson, Charlie Hoy, Anna Heffernan, Meryl Kinnear, Jorge Valencia, Felip A Ramis Vidal, Cecilio Garc\'ia-Quir\'os, Shrobana Ghosgh, Lionel London, Mark Hannam and Sascha Husa | PhenomXPNR: An improved gravitational wave model linking precessing inspirals and NR-calibrated merger-ringdown | 18 pages, 16 figures | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | We present the frequency-domain quasi-circular precessing binary-black-hole model PhenomXPNR. This model combines the most precise available post-Newtonian description of the evolution of the precession dynamics through inspiral with merger-ringdown model informed by numerical relativity. This, along with a phenomenological model of the dominant multipole asymmetries, results in the most accurate and complete representation of the physics of precessing binaries natively in the frequency-domain to date. All state-of-the-art precessing models show bias when inferring binary parameters in certain regions of the parameter space. We demonstrate that the developments presented ensure that for some precessing systems PhenomXPNR shows the least degree of bias. Further, as a phenomenological, frequency-domain model, PhenomXPNR remains one of the most computationally efficient models available and is therefore well-suited to the era of gravitational-wave astronomy with its ever growing rate of detected signals. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:25:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Hamilton",
"Eleanor",
""
],
[
"Colleoni",
"Marta",
""
],
[
"Thompson",
"Jonathan E.",
""
],
[
"Hoy",
"Charlie",
""
],
[
"Heffernan",
"Anna",
""
],
[
"Kinnear",
"Meryl",
""
],
[
"Valencia",
"Jorge",
""
],
[
"Vidal",
"Felip A Ramis",
""
],
[
"García-Quirós",
"Cecilio",
""
],
[
"Ghosgh",
"Shrobana",
""
],
[
"London",
"Lionel",
""
],
[
"Hannam",
"Mark",
""
],
[
"Husa",
"Sascha",
""
]
] |
2507.02605 | Yizhuang Liu | Yizhuang Liu, Marcos Mari\~no | Trans-series from condensates in the non-linear sigma model | 39 pages, 8 figures | null | null | null | hep-th hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work we provide a massless perturbative framework for the two dimensional non-linear sigma model (NLSM), that allows the computation of the perturbative series attached to the operator condensates in the operator product expansion (OPE). It is based on a limit of the quartic linear sigma model (LSM) and is manifestly $O(N)$ symmetric. We show, at next-to-leading order in the $1/N$ expansion, how this framework reproduces the perturbative contribution to the two-point function, as well as its first exponentially small correction due to the condensate of the Lagrangian operator, in full agreement with the exact non-perturbative large $N$ solution. We also show that, in the full LSM, the physics at the natural UV cutoff indeed decouples from the NLSM in the IR, in the weak-coupling limit. In particular, we show that the perturbative framework for the LSM at the cutoff scale is connected to the one in the NLSM. The structure of power divergences in the LSM regularization also reveals that the first renormalon on the positive Borel axis of the NLSM perturbative self-energy is an UV renormalon, which cancels against the ambiguity in the condensate. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:27:25 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Yizhuang",
""
],
[
"Mariño",
"Marcos",
""
]
] |
2507.02606 | Wei Fan | Wei Fan, Kejiang Chen, Chang Liu, Weiming Zhang, Nenghai Yu | De-AntiFake: Rethinking the Protective Perturbations Against Voice Cloning Attacks | Accepted by ICML 2025 | null | null | null | cs.SD cs.AI cs.CR cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid advancement of speech generation models has heightened privacy and security concerns related to voice cloning (VC). Recent studies have investigated disrupting unauthorized voice cloning by introducing adversarial perturbations. However, determined attackers can mitigate these protective perturbations and successfully execute VC. In this study, we conduct the first systematic evaluation of these protective perturbations against VC under realistic threat models that include perturbation purification. Our findings reveal that while existing purification methods can neutralize a considerable portion of the protective perturbations, they still lead to distortions in the feature space of VC models, which degrades the performance of VC. From this perspective, we propose a novel two-stage purification method: (1) Purify the perturbed speech; (2) Refine it using phoneme guidance to align it with the clean speech distribution. Experimental results demonstrate that our method outperforms state-of-the-art purification methods in disrupting VC defenses. Our study reveals the limitations of adversarial perturbation-based VC defenses and underscores the urgent need for more robust solutions to mitigate the security and privacy risks posed by VC. The code and audio samples are available at https://de-antifake.github.io. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:30:58 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fan",
"Wei",
""
],
[
"Chen",
"Kejiang",
""
],
[
"Liu",
"Chang",
""
],
[
"Zhang",
"Weiming",
""
],
[
"Yu",
"Nenghai",
""
]
] |
2507.02607 | Mahshid Helali Moghadam | Frida Sundfeldt, Bianca Widstam, Mahshid Helali Moghadam, Kuo-Yun Liang, Anders Vesterberg | Alleviating Attack Data Scarcity: SCANIA's Experience Towards Enhancing In-Vehicle Cyber Security Measures | null | null | null | null | cs.CR cs.LG cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The digital evolution of connected vehicles and the subsequent security risks emphasize the critical need for implementing in-vehicle cyber security measures such as intrusion detection and response systems. The continuous advancement of attack scenarios further highlights the need for adaptive detection mechanisms that can detect evolving, unknown, and complex threats. The effective use of ML-driven techniques can help address this challenge. However, constraints on implementing diverse attack scenarios on test vehicles due to safety, cost, and ethical considerations result in a scarcity of data representing attack scenarios. This limitation necessitates alternative efficient and effective methods for generating high-quality attack-representing data. This paper presents a context-aware attack data generator that generates attack inputs and corresponding in-vehicle network log, i.e., controller area network (CAN) log, representing various types of attack including denial of service (DoS), fuzzy, spoofing, suspension, and replay attacks. It utilizes parameterized attack models augmented with CAN message decoding and attack intensity adjustments to configure the attack scenarios with high similarity to real-world scenarios and promote variability. We evaluate the practicality of the generated attack-representing data within an intrusion detection system (IDS) case study, in which we develop and perform an empirical evaluation of two deep neural network IDS models using the generated data. In addition to the efficiency and scalability of the approach, the performance results of IDS models, high detection and classification capabilities, validate the consistency and effectiveness of the generated data as well. In this experience study, we also elaborate on the aspects influencing the fidelity of the data to real-world scenarios and provide insights into its application. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:31:33 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Sundfeldt",
"Frida",
""
],
[
"Widstam",
"Bianca",
""
],
[
"Moghadam",
"Mahshid Helali",
""
],
[
"Liang",
"Kuo-Yun",
""
],
[
"Vesterberg",
"Anders",
""
]
] |
2507.02608 | Fran\c{c}ois Rozet | Fran\c{c}ois Rozet, Ruben Ohana, Michael McCabe, Gilles Louppe, Fran\c{c}ois Lanusse, Shirley Ho | Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulation | null | null | null | null | cs.LG physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The steep computational cost of diffusion models at inference hinders their use as fast physics emulators. In the context of image and video generation, this computational drawback has been addressed by generating in the latent space of an autoencoder instead of the pixel space. In this work, we investigate whether a similar strategy can be effectively applied to the emulation of dynamical systems and at what cost. We find that the accuracy of latent-space emulation is surprisingly robust to a wide range of compression rates (up to 1000x). We also show that diffusion-based emulators are consistently more accurate than non-generative counterparts and compensate for uncertainty in their predictions with greater diversity. Finally, we cover practical design choices, spanning from architectures to optimizers, that we found critical to train latent-space emulators. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:32:50 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Rozet",
"François",
""
],
[
"Ohana",
"Ruben",
""
],
[
"McCabe",
"Michael",
""
],
[
"Louppe",
"Gilles",
""
],
[
"Lanusse",
"François",
""
],
[
"Ho",
"Shirley",
""
]
] |
2507.02609 | Vitaly Khonik Alexandrovich | A.S. Makarov, R.A. Konchakov, N.P. Kobelev, V.A. Khonik | Thermodynamic potentials of metallic alloys in the undercooled liquid and solid glassy states | 18 pages, 8 figures | null | null | null | cond-mat.dis-nn | http://creativecommons.org/licenses/by/4.0/ | We first present a comparative analysis of temperature evolution of the excess thermodynamic potentials (state functions), the enthalpy $\Delta H$, entropy $\Delta S$ and Gibbs free energy $\Delta \Phi$, determined for \textit{i}) undercooled melts using literature data and \textit{ii}) solid glassy state calculated on the basis of calorimetry measurements using an approach proposed recently. Three metallic alloys were taken as an example for data analysis. It is found that temperature dependences $\Delta H(T)$, $\Delta S(T)$ and $\Delta G(T)$ calculated with both approaches coincide in the supercooled liquid range (i.e. at temperatures $T_g<T<T_x$, where $T_g$ and $T_x$ are the glass transition and crystallization onset temperatures, respectively). However, the necessary conditions for this coincidence is the introduction of important changes to the above approach \textit{i}), which are related to the calculation of the melting entropy. We also introduce and calculate a dimensionless order parameter $\xi$, which changes in the range $0<\xi<1$ and characterizes the evolution of the structural order from liquid-like to crystal-like one. It is shown that the order parameter $\xi_{scl}$ calculated for the end of the supercooled liquid range (i.e. for a temperature just below $T_x$) correlates with the melt critical cooling rate $R_c$: the smaller the order parameter $\xi_{scl}$ (i.e. the closer the structure to that of the equilibrium liquid), the smaller $R_c$ is. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:33:01 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Makarov",
"A. S.",
""
],
[
"Konchakov",
"R. A.",
""
],
[
"Kobelev",
"N. P.",
""
],
[
"Khonik",
"V. A.",
""
]
] |
2507.02610 | Abdelhakim Benkrane | Abdelhakim Benkrane | Entropy Analysis of Dark Matter Halo Structures | null | null | null | null | gr-qc | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this study, we aim to derive the entropy associated with a dark matter halo modeled using a double (broken) power-law density profile. Our approach is inspired by the pioneering work of Verlinde, who proposed that gravity may not be a fundamental force but rather an emergent phenomenon rooted in entropic principles. To investigate this idea, we examine four different dark matter halo profiles: the Dehnen-type, Hernquist, Jaffe, Plummer sphere, and the perfect sphere. Using these profiles, we derive corresponding entropy expressions and assess their consistency with the second law of thermodynamics, which governs the behavior of entropy in closed systems. Furthermore, to broaden the scope of our analysis, we extend the study to a cosmological context, allowing us to explore how the derived entropy expressions influence the dynamics of the Friedmann equations that describe the evolution of the universe. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:33:57 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Benkrane",
"Abdelhakim",
""
]
] |
2507.02611 | Kai-Ge Zhang | Kai-Ge Zhang, Jian-Feng He, Chengjie Fu, Zong-Kuan Guo | Gravitational Waves from Gauge Quanta Produced during Inflation | null | null | null | null | hep-ph astro-ph.CO gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A fast-rolling axion can transfer its kinetic energy to a gauge field via the Chern-Simons coupling, leading to copious production of gauge quanta, which can act as a source of gravitational waves (GWs) with potentially observable amplitudes. In this work, we investigate GW production in a spectator axion model when strong backreaction is taken into account. We find that decreasing the decay constant of the axion enhances GW production. Since the initial value of the axion is larger than its quantum fluctuations, such a condition imposes a lower bound on the axion dacay constant, which sets an upper bound on the amplitude of the energy spectrum of GWs. As a result, the amplitude of the predicted GW energy spectrum is lower than $10^{-10}$ in the nHz to mHz frequency range. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:34:36 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Kai-Ge",
""
],
[
"He",
"Jian-Feng",
""
],
[
"Fu",
"Chengjie",
""
],
[
"Guo",
"Zong-Kuan",
""
]
] |
2507.02612 | Maximilian Daschner | Maximilian Daschner, Ivan Kokanovi\'c and F. Malte Grosche | Mechanical enhancement of quantum oscillations | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | We investigate quantum oscillation measurements in the Dirac nodal-line semimetal TaNiTe$_5$ which exhibit a strongly enhanced amplitude in the magnetoresistance. We show that mechanical properties of the measurement setup in combination with de Haas - van Alphen oscillations in the magnetic torque can cause this enhancement in the measured resistance, without involvement of any topological properties in this material. To support the empirical data, a numerical model is provided, showing good agreement. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:36:43 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Daschner",
"Maximilian",
""
],
[
"Kokanović",
"Ivan",
""
],
[
"Grosche",
"F. Malte",
""
]
] |
2507.02613 | Tugba Erpek | Yalin E. Sagduyu, Kemal Davaslioglu, Tugba Erpek, Sastry Kompella, Gustave Anderson, Jonathan Ashdown | MULTI-SCOUT: Multistatic Integrated Sensing and Communications in 5G and Beyond for Moving Target Detection, Positioning, and Tracking | null | null | null | null | cs.NI cs.DC eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a complete signal-processing chain for multistatic integrated sensing and communications (ISAC) using 5G Positioning Reference Signal (PRS). We consider a distributed architecture in which one gNB transmits a periodic OFDM-PRS waveform while multiple spatially separated receivers exploit the same signal for target detection, parameter estimation and tracking. A coherent cross-ambiguity function (CAF) is evaluated to form a range-Doppler map from which the bistatic delay and radial velocity are extracted for every target. For a single target, the resulting bistatic delays are fused through nonlinear least-squares trilateration, yielding a geometric position estimate, and a regularized linear inversion of the radial-speed equations yields a two-dimensional velocity vector, where speed and heading are obtained. The approach is applied to 2D and 3D settings, extended to account for time synchronization bias, and generalized to multiple targets by resolving target association. The sequence of position-velocity estimates is then fed to standard and extended Kalman filters to obtain smoothed tracks. Our results show high-fidelity moving-target detection, positioning, and tracking using 5G PRS signals for multistatic ISAC. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:38:19 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Sagduyu",
"Yalin E.",
""
],
[
"Davaslioglu",
"Kemal",
""
],
[
"Erpek",
"Tugba",
""
],
[
"Kompella",
"Sastry",
""
],
[
"Anderson",
"Gustave",
""
],
[
"Ashdown",
"Jonathan",
""
]
] |
2507.02614 | Diogo Sousa | Diogo Sousa | A Dopamine-Serotonin Theory of Consciousness | 37 pages, 16 figures, 3 tables. Includes extensive theoretical modeling, data analysis of 70,290 patient-nights, and supplementary sections on developmental trajectories and future therapeutic protocols. All analysis code is available at: https://github.com/DiogoSousa509/dopa-5ht-consciousness | null | null | null | q-bio.NC q-bio.OT | http://creativecommons.org/licenses/by/4.0/ | This work presents a comprehensive theory of consciousness grounded in mathematical formalism and supported by clinical data analysis. The framework developed herein demonstrates that consciousness exists as a continuous, non-monotonic function across a high-dimensional neurochemical space, with dopamine serving as the primary intensity regulator and serotonin (5-HT2A) as the complexity modulator. This work offers mechanistic explanations for the full spectrum of conscious states, from deep sleep and psychosis to the ultimate collapse in neural death. The theory explains paradoxical phenomena such as prefrontal cortex hypoactivity during seizures, the evolutionary persistence of psychosis-prone individuals, and why controlled administration of classical 5-HT2A agonists shows a comparatively low incidence of serious medical events (< 0.01 % in modern clinical trials), while dopaminergic excess proves rapidly lethal. The framework is tested using 70,290 sleep nights from 242 Parkinson's disease patients, using disease severity (UPDRS) as a proxy for system integrity and medication (LEDD) as a proxy for dopaminergic input. The analysis reveals a significant LEDD x UPDRS interaction (beta=-1.7, p<.0001), confirming the model's prediction of state-dependent, non-linear dynamics. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:38:40 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Sousa",
"Diogo",
""
]
] |
2507.02615 | Hanshang Jin | Hanshang Jin, Rahim R. Ullah, Peter Klavins, Valentin Taufour | Importance of anisotropic interactions for hard-axis/plane ordering of Ce-based ferromagnets | Accepted by PRB | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Ferromagnetic (FM) Kondo-lattice (KL) compounds often exhibit intriguing magnetic behavior driven by strong crystal electric field (CEF) anisotropy and exchange interactions. Recent studies suggest that many Ce-based and Yb-based KL ferromagnets order along the hard-axis or hard-plane determined by their CEF ground state. We performed a survey of Ce-based FM compounds, complemented with new single crystal synthesis, magnetization measurements, and analysis of the CEF scheme. Our results reveal that hard-axis/plane ordering is less common than previously reported, with most compounds ordering along their easy-axis/plane. We also find no clear correlation between the strength of the Kondo effect and whether a compound adopts hard-axis/plane or easy-axis/plane ordering. Instead, the key driver appears to be the competition between CEF anisotropy and exchange interaction anisotropy. A direct comparison of CeCuSi and CeAgSb$_2$ further indicates that antiferromagnetic (AFM) interactions along the CEF easy-axis can be crucial in stabilizing the hard-axis FM ordering observed in CeAgSb$_2$. Finally, our results suggest that an anisotropic Ruderman-Kittel-Kasuya-Yosida (RKKY) model could offer deeper insights into the complicated magnetic properties of Ce-based and other heavy-fermion systems. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:42:57 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Jin",
"Hanshang",
""
],
[
"Ullah",
"Rahim R.",
""
],
[
"Klavins",
"Peter",
""
],
[
"Taufour",
"Valentin",
""
]
] |
2507.02616 | Tianqi Shang | Tianqi Shang, Weiqing He, Charles Zheng, Lingyao Li, Li Shen, Bingxin Zhao | DynamiCare: A Dynamic Multi-Agent Framework for Interactive and Open-Ended Medical Decision-Making | 16 pages | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The rise of Large Language Models (LLMs) has enabled the development of specialized AI agents with domain-specific reasoning and interaction capabilities, particularly in healthcare. While recent frameworks simulate medical decision-making, they largely focus on single-turn tasks where a doctor agent receives full case information upfront -- diverging from the real-world diagnostic process, which is inherently uncertain, interactive, and iterative. In this paper, we introduce MIMIC-Patient, a structured dataset built from the MIMIC-III electronic health records (EHRs), designed to support dynamic, patient-level simulations. Building on this, we propose DynamiCare, a novel dynamic multi-agent framework that models clinical diagnosis as a multi-round, interactive loop, where a team of specialist agents iteratively queries the patient system, integrates new information, and dynamically adapts its composition and strategy. We demonstrate the feasibility and effectiveness of DynamiCare through extensive experiments, establishing the first benchmark for dynamic clinical decision-making with LLM-powered agents. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:43:10 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shang",
"Tianqi",
""
],
[
"He",
"Weiqing",
""
],
[
"Zheng",
"Charles",
""
],
[
"Li",
"Lingyao",
""
],
[
"Shen",
"Li",
""
],
[
"Zhao",
"Bingxin",
""
]
] |
2507.02617 | Xian Dai | Xian Dai, Gerhard Knieper | Maximal stretch and Lipschitz maps on Riemannian manifolds of negative curvature | 83 pages. Comments welcome | null | null | null | math.DG math.DS math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In his seminal work on Teichm\"uller spaces (\cite{Th98}), Thurston introduced the maximal stretch for a pair of hyperbolic metrics on a closed surface of genus $\geq 2$ and showed that the logarithm of this quantity induces an asymmetric metric in the Teichm\"uller space. He also showed that the subset of the surface on which the maximal stretch is attained is a geodesic lamination. In this paper, we define the maximal stretch analogously for closed manifolds equipped with Riemannian metrics of variable negative curvature and investigate the structure of the related Mather set on the unit tangent bundle. In contrast to the Teichm\"uller space, the Mather set may not be lifts of geodesic laminations in this broader setting. However, in our paper, we will discuss similar features shared by the Mather set with geodesic laminations. We also connect the study of the Mather set with the theory of best Lipschitz maps. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:44:22 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Dai",
"Xian",
""
],
[
"Knieper",
"Gerhard",
""
]
] |
2507.02618 | Kenneth Payne | Kenneth Payne and Baptiste Alloui-Cros | Strategic Intelligence in Large Language Models: Evidence from evolutionary Game Theory | 29 pages, 27 tables, 4 figures | null | null | null | cs.AI cs.CL cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Are Large Language Models (LLMs) a new form of strategic intelligence, able to reason about goals in competitive settings? We present compelling supporting evidence. The Iterated Prisoner's Dilemma (IPD) has long served as a model for studying decision-making. We conduct the first ever series of evolutionary IPD tournaments, pitting canonical strategies (e.g., Tit-for-Tat, Grim Trigger) against agents from the leading frontier AI companies OpenAI, Google, and Anthropic. By varying the termination probability in each tournament (the "shadow of the future"), we introduce complexity and chance, confounding memorisation.
Our results show that LLMs are highly competitive, consistently surviving and sometimes even proliferating in these complex ecosystems. Furthermore, they exhibit distinctive and persistent "strategic fingerprints": Google's Gemini models proved strategically ruthless, exploiting cooperative opponents and retaliating against defectors, while OpenAI's models remained highly cooperative, a trait that proved catastrophic in hostile environments. Anthropic's Claude emerged as the most forgiving reciprocator, showing remarkable willingness to restore cooperation even after being exploited or successfully defecting. Analysis of nearly 32,000 prose rationales provided by the models reveals that they actively reason about both the time horizon and their opponent's likely strategy, and we demonstrate that this reasoning is instrumental to their decisions. This work connects classic game theory with machine psychology, offering a rich and granular view of algorithmic decision-making under uncertainty. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:45:02 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Payne",
"Kenneth",
""
],
[
"Alloui-Cros",
"Baptiste",
""
]
] |
2507.02619 | Hazal Mogultay Ozcan | Hazal Mogultay Ozcan, Sinan Kalkan, Fatos T. Yarman-Vural | L-VAE: Variational Auto-Encoder with Learnable Beta for Disentangled Representation | The paper is under revision at Machine Vision and Applications | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a novel model called Learnable VAE (L-VAE), which learns a disentangled representation together with the hyperparameters of the cost function. L-VAE can be considered as an extension of \b{eta}-VAE, wherein the hyperparameter, \b{eta}, is empirically adjusted. L-VAE mitigates the limitations of \b{eta}-VAE by learning the relative weights of the terms in the loss function to control the dynamic trade-off between disentanglement and reconstruction losses. In the proposed model, the weight of the loss terms and the parameters of the model architecture are learned concurrently. An additional regularization term is added to the loss function to prevent bias towards either reconstruction or disentanglement losses. Experimental analyses show that the proposed L-VAE finds an effective balance between reconstruction fidelity and disentangling the latent dimensions. Comparisons of the proposed L-VAE against \b{eta}-VAE, VAE, ControlVAE, DynamicVAE, and {\sigma}-VAE on datasets, such as dSprites, MPI3D-complex, Falcor3D, and Isaac3D reveals that L-VAE consistently provides the best or the second best performances measured by a set of disentanglement metrics. Moreover, qualitative experiments on CelebA dataset, confirm the success of the L-VAE model for disentangling the facial attributes. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:45:42 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ozcan",
"Hazal Mogultay",
""
],
[
"Kalkan",
"Sinan",
""
],
[
"Yarman-Vural",
"Fatos T.",
""
]
] |
2507.02620 | Xing Liu | Xing Liu, Lizhuo Luo, Ming Tang, Chao Huang | FlowSpec: Continuous Pipelined Speculative Decoding for Efficient Distributed LLM Inference | 16 pages, and the last 3 are appendix | null | null | null | cs.DC cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Distributed inference serves as a promising approach to enabling the inference of large language models (LLMs) at the network edge. It distributes the inference process to multiple devices to ensure that the LLMs can fit into the device memory. Recent pipeline-based approaches have the potential to parallelize communication and computation, which helps reduce inference latency. However, the benefit diminishes when the inference request at the network edge is sparse, where pipeline is typically at low utilization. To enable efficient distributed LLM inference at the edge, we propose \textbf{FlowSpec}, a pipeline-parallel tree-based speculative decoding framework. FlowSpec incorporates three key mechanisms to improve decoding efficiency: 1) score-based step-wise verification prioritizes more important draft tokens to bring earlier accpeted tokens; 2) efficient draft management to prune invalid tokens while maintaining correct causal relationship during verification; 3) dynamic draft expansion strategies to supply high-quality speculative inputs. These techniques work in concert to enhance both pipeline utilization and speculative efficiency. We evaluate FlowSpec on a real-world testbed with other baselines. Experimental results demonstrate that our proposed framework significantly improves inference speed across diverse models and configurations, achieving speedup ratios 1.36$\times$-1.77$\times$ compared to baselines. Our code is publicly available at \href{https://github.com/Leosang-lx/FlowSpec#}{https://github.com/Leosang-lx/FlowSpec\#} | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:47:42 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Xing",
""
],
[
"Luo",
"Lizhuo",
""
],
[
"Tang",
"Ming",
""
],
[
"Huang",
"Chao",
""
]
] |
2507.02621 | Cheng-Liang Jiao | Cheng-Liang Jiao, Er-gang Zhao, Liying Zhu and Xiang-dong Shi | Spherically Symmetric Accretion with Self-Gravity: Analytical Formulae and Numerical Validation | 16 pages, 3 figures, accepted for publication in ApJ | null | null | null | astro-ph.HE astro-ph.GA astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spherically symmetric accretion incorporating self-gravity constitutes a three-point boundary value problem (TPBVP) governed by constraints at the outer boundary, sonic point, and accretor surface. Previous studies have two limitations: either employing an incorrect formula for self-gravity potential in analytical treatments, or introducing additional input parameters in numerical implementations to circumvent solving the full TPBVP. To address these issues, we present a self-consistent TPBVP formulation, solved using the relaxation method. We also derive approximate analytical formulae that enable rapid estimates of self-gravity effects. Our analysis identifies a dimensionless parameter $\beta \equiv 2G \bar{\rho} r_\mathrm{out}^2/a_\mathrm{out}^2$ that characterizes the strength of self-gravity, where $\bar{\rho}$ and $r_\mathrm{out}$ are the mean density and outer radius of the flow, respectively, and $a_\mathrm{out}$ is the adiabatic sound speed of the external medium. For practical estimation, $\bar{\rho}$ may be approximated by the external medium density $\rho_\mathrm{out}$. We identify an upper limit for $\beta$, beyond which steady accretion becomes unsustainable -- a behavior consistent with classical gravitational instability that previous studies failed to capture. The accretion rate enhancement decreases monotonically as the adiabatic index $\gamma$ increases. For $\gamma=5/3$, self-gravity ceases to augment the accretion rate. These theoretical predictions are validated by our numerical solutions. We further apply our results to two astrophysical scenarios: hyper-Eddington accretion onto supermassive black hole seeds in the early Universe, where self-gravity is significant; and accretion onto stellar-mass objects embedded in active galactic nuclei (AGN) disks, where self-gravity is non-negligible under certain conditions and should be evaluated using $\beta$. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:47:49 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Jiao",
"Cheng-Liang",
""
],
[
"Zhao",
"Er-gang",
""
],
[
"Zhu",
"Liying",
""
],
[
"Shi",
"Xiang-dong",
""
]
] |
2507.02622 | Zhicheng Zhang | Zhicheng Zhang and Mingsheng Ying | Access Control Threatened by Quantum Entanglement | 23 pages, 10 figures | null | null | null | quant-ph cs.CR cs.OS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Access control is a cornerstone of computer security that prevents unauthorised access to resources. In this paper, we study access control in quantum computer systems. We present the first explicit scenario of a security breach when a classically secure access control system is straightforwardly adapted to the quantum setting. The breach is ultimately due to that quantum mechanics allows the phenomenon of entanglement and violates Mermin inequality, a multi-party variant of the celebrated Bell inequality. This reveals a threat from quantum entanglement to access control if existing computer systems integrate with quantum computing. To protect against such threat, we propose several new models of quantum access control, and rigorously analyse their security, flexibility and efficiency. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:48:14 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Zhicheng",
""
],
[
"Ying",
"Mingsheng",
""
]
] |
2507.02623 | Lucas Ramos | L. M. Ramos, F. M. Zimmer, M. Schmidt | Interplay of frustration and quantum fluctuations in a spin-1/2 anisotropic square lattice | 9 pages, 8 figures | Phys. Rev. B 112, 014402, 2025 | 10.1103/svsc-bvjx | null | cond-mat.str-el cond-mat.mtrl-sci cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | Motivated by theoretical and experimental studies reported by Yamaguchi et al. (Phys. Rev. B 98, 094402 (2018)), we performed a cluster mean-field analysis of an anisotropic Heisenberg model with six competing exchange interactions. We study the ground and thermal states by tuning the spin anisotropy, magnetic field, and temperature. Our results show that an external magnetic field induces quantum fluctuations, suppressing local moments and leading to the occurrence of a magnetization plateaulike state. When weak spin anisotropy is considered, the competing interactions are affected, and the field-induced fluctuations can lead to a well-defined magnetization plateau within a field range, in which an exotic quantum state can emerge. This state exhibits the coexistence of ferromagnetic and dimerized chains driven by the relation between frustration and the external field. Moreover, we identify a phase transition from a collinear antiferromagnetic order to a disordered state at a finite temperature. Our findings reveal unconventional magnetic properties at low temperatures that can guide future experimental studies of verdazyl-based compounds with anisotropic exchange interactions. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:50:16 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ramos",
"L. M.",
""
],
[
"Zimmer",
"F. M.",
""
],
[
"Schmidt",
"M.",
""
]
] |
2507.02624 | Antoine Honor\'e | Antoine Honor\'e, Borja Rodr\'iguez G\'alvez, Yoomi Park, Yitian Zhou, Volker M. Lauschke, Ming Xiao | A Matrix Variational Auto-Encoder for Variant Effect Prediction in Pharmacogenes | 12+8 pages | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Variant effect predictors (VEPs) aim to assess the functional impact of protein variants, traditionally relying on multiple sequence alignments (MSAs). This approach assumes that naturally occurring variants are fit, an assumption challenged by pharmacogenomics, where some pharmacogenes experience low evolutionary pressure. Deep mutational scanning (DMS) datasets provide an alternative by offering quantitative fitness scores for variants. In this work, we propose a transformer-based matrix variational auto-encoder (matVAE) with a structured prior and evaluate its performance on 33 DMS datasets corresponding to 26 drug target and ADME proteins from the ProteinGym benchmark. Our model trained on MSAs (matVAE-MSA) outperforms the state-of-the-art DeepSequence model in zero-shot prediction on DMS datasets, despite using an order of magnitude fewer parameters and requiring less computation at inference time. We also compare matVAE-MSA to matENC-DMS, a model of similar capacity trained on DMS data, and find that the latter performs better on supervised prediction tasks. Additionally, incorporating AlphaFold-generated structures into our transformer model further improves performance, achieving results comparable to DeepSequence trained on MSAs and finetuned on DMS. These findings highlight the potential of DMS datasets to replace MSAs without significant loss in predictive performance, motivating further development of DMS datasets and exploration of their relationships to enhance variant effect prediction. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:50:18 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Honoré",
"Antoine",
""
],
[
"Gálvez",
"Borja Rodríguez",
""
],
[
"Park",
"Yoomi",
""
],
[
"Zhou",
"Yitian",
""
],
[
"Lauschke",
"Volker M.",
""
],
[
"Xiao",
"Ming",
""
]
] |
2507.02625 | Serge Florens | Izak Snyman, Serge Florens | The covariance matrix spectrum of correlated charge insulators reveals hidden connections to Coupled Cluster, Matrix Product, and Rokhsar-Kivelson states | 14+11 pages, 7 figures | null | null | null | cond-mat.str-el cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | Charge ordering induced by strong short-range repulsion in itinerant fermion systems typically follows a two-sites alternation pattern. However, the covariance matrix spectrum of the one-dimensional, half-filled, spinless $t$-$V$ model reveals a post-Hartree-Fock picture at strong repulsion, with emergent four-site disruptions of the underlying staggered mean-field state. These disruptions are captured in a thermodynamically extensive manner by a compact four-fermion Coupled Cluster (doubles) state (CCS). Remarkably, all properties of this state may be computed analytically by combinatorial means, and also derived from an exactly solvable correlated hopping Hamiltonian. Furthermore, this Coupled Cluster state can be re-expressed as a low-rank Matrix Product State (MPS) with bond dimension exactly four. In addition, we unveil a hidden connection between this Coupled Cluster ansatz and a Rokhsar-Kivelson state (RKS), which is the ground state of a solvable parent quantum tetramer model. The broad picture that we uncover here thus provides deep connections between several core concepts of correlated fermions and quantum chemistry that have previously enjoyed limited synergy. In contrast to a recent perturbative treatment on top of Hartree-Fock theory, our approach asymptotically captures the correct correlations in the $t$-$V$ model at small $t/V$, and remains a qualitatively accurate approximation even outside the perturbative regime. Our results make the case for further studies of the covariance matrix for correlated electron systems in which ground states have non-trivial unit-cell structure. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:50:40 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Snyman",
"Izak",
""
],
[
"Florens",
"Serge",
""
]
] |
2507.02626 | Siran Chen | Siran Chen, Boyu Chen, Chenyun Yu, Yuxiao Luo, Ouyang Yi, Lei Cheng, Chengxiang Zhuo, Zang Li, Yali Wang | VRAgent-R1: Boosting Video Recommendation with MLLM-based Agents via Reinforcement Learning | null | null | null | null | cs.MM | http://creativecommons.org/licenses/by/4.0/ | Owing to powerful natural language processing and generative capabilities, large language model (LLM) agents have emerged as a promising solution for enhancing recommendation systems via user simulation. However, in the realm of video recommendation, existing studies predominantly resort to prompt-based simulation using frozen LLMs and encounter the intricate challenge of multimodal content understanding. This frequently results in suboptimal item modeling and user preference learning, thereby ultimately constraining recommendation performance. To address these challenges, we introduce VRAgent-R1, a novel agent-based paradigm that incorporates human-like intelligence in user simulation. Specifically, VRAgent-R1 comprises two distinct agents: the Item Perception (IP) Agent and the User Simulation (US) Agent, designed for interactive user-item modeling. Firstly, the IP Agent emulates human-like progressive thinking based on MLLMs, effectively capturing hidden recommendation semantics in videos. With a more comprehensive multimodal content understanding provided by the IP Agent, the video recommendation system is equipped to provide higher-quality candidate items. Subsequently, the US Agent refines the recommended video sets based on in-depth chain-of-thought (CoT) reasoning and achieves better alignment with real user preferences through reinforcement learning. Experimental results on a large-scale video recommendation benchmark have demonstrated the effectiveness of our proposed VRAgent-R1 method, e.g., the IP Agent achieves a 6.0\% improvement in NDCG@10 on the MicroLens-100k dataset, while the US Agent shows approximately 45.0\% higher accuracy in user decision simulation compared to state-of-the-art baselines. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:52:24 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Chen",
"Siran",
""
],
[
"Chen",
"Boyu",
""
],
[
"Yu",
"Chenyun",
""
],
[
"Luo",
"Yuxiao",
""
],
[
"Yi",
"Ouyang",
""
],
[
"Cheng",
"Lei",
""
],
[
"Zhuo",
"Chengxiang",
""
],
[
"Li",
"Zang",
""
],
[
"Wang",
"Yali",
""
]
] |
2507.02627 | Frank den Hollander | Bishakh Bhattacharya, Nitya Gadhiwala, Frank den Hollander, Pradeeptha R Jain, Tejashree Subramanya | The Triangle Friendship Paradox | 29 pages, 9 figures, 2 tables | null | null | null | math.PR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We consider the generalised friendship paradox, focussing on the number of triangles at a vertex as the relevant attribute. We show that, contrary to the setting where the attribute is the number of edges at a vertex or the number of wedges at a vertex, the average friendship-bias of the number of triangles at a vertex is not always non-negative. We identify classes of finite deterministic graphs for which the bias is non-negative, and provide examples of finite deterministic graphs for which it is not. For certain classes of sparse and dense random graphs, we compute the scaling of the bias in the limit as the number of vertices tends to infinity. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:53:20 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Bhattacharya",
"Bishakh",
""
],
[
"Gadhiwala",
"Nitya",
""
],
[
"Hollander",
"Frank den",
""
],
[
"Jain",
"Pradeeptha R",
""
],
[
"Subramanya",
"Tejashree",
""
]
] |
2507.02628 | Irena Girshovitz | Irena Girshovitz, Atai Ambus, Moni Shahar, Ran Gilad-Bachrach | Medical Data Pecking: A Context-Aware Approach for Automated Quality Evaluation of Structured Medical Data | 18 pages, 4 figures (+ appendix) | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Background: The use of Electronic Health Records (EHRs) for epidemiological studies and artificial intelligence (AI) training is increasing rapidly. The reliability of the results depends on the accuracy and completeness of EHR data. However, EHR data often contain significant quality issues, including misrepresentations of subpopulations, biases, and systematic errors, as they are primarily collected for clinical and billing purposes. Existing quality assessment methods remain insufficient, lacking systematic procedures to assess data fitness for research.
Methods: We present the Medical Data Pecking approach, which adapts unit testing and coverage concepts from software engineering to identify data quality concerns. We demonstrate our approach using the Medical Data Pecking Tool (MDPT), which consists of two main components: (1) an automated test generator that uses large language models and grounding techniques to create a test suite from data and study descriptions, and (2) a data testing framework that executes these tests, reporting potential errors and coverage.
Results: We evaluated MDPT on three datasets: All of Us (AoU), MIMIC-III, and SyntheticMass, generating 55-73 tests per cohort across four conditions. These tests correctly identified 20-43 non-aligned or non-conforming data issues. We present a detailed analysis of the LLM-generated test suites in terms of reference grounding and value accuracy.
Conclusion: Our approach incorporates external medical knowledge to enable context-sensitive data quality testing as part of the data analysis workflow to improve the validity of its outcomes. Our approach tackles these challenges from a quality assurance perspective, laying the foundation for further development such as additional data modalities and improved grounding methods. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:54:50 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Girshovitz",
"Irena",
""
],
[
"Ambus",
"Atai",
""
],
[
"Shahar",
"Moni",
""
],
[
"Gilad-Bachrach",
"Ran",
""
]
] |
2507.02629 | Igor Fedorov | Igor A. Fedorov | Pressure-induced band gap energy increase in crystalline lactose | null | null | 10.1016/j.jpcs.2025.112991 | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Lactose is widely used in agri-food and pharma industries. New materials that can be obtained from renewable sources are currently being sought. I have studied the effect of hydrostatic pressure on structural properties of currently known forms of crystalline lactose within the framework of density functional theory with van der Waals interactions. The computed parameters have good agreement with experimental data. The effect of mechanical deformations on electron structure of crystalline lactose was also studied. Compression of the crystal leads to the band gap increase. The analysis of partial density of states of lactose crystals was performed. The band gap of different forms of crystalline lactose was also computed using the quasiparticle G0W0 approximation. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:56:12 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fedorov",
"Igor A.",
""
]
] |
2507.02630 | Anshu Kumari | Srinjana Routh, Anshu Kumari, Vaibhav Pant, Jaydeep Kandekar, Dipankar Banerjee, Mohd. Saleem Khan, and Dibya Kirti Mishra | Insights into Chromospheric Large-Scale Flows using Nobeyama 17 GHz Radio Observations I. The Differential Rotation Profile | 10 Pages, 6 figures, 2 tables, Accepted for publication in Astronomy and Astrophysics Letters | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Although the differential rotation rate on the solar surface has long been studied using optical and extreme ultraviolet (EUV) observations, associating these measurements to specific atmospheric heights remains challenging due to the temperature-dependent emission of tracers observed in EUV wavelengths. Radio observations, being primarily influenced by coherent plasma processes and/or thermal bremsstrahlung, offer a more height-stable diagnostic and thus provide an independent means to test and validate rotational trends observed at other EUV wavelengths. We aim to characterize the differential rotation profile of the upper chromosphere using cleaned solar full-disc 17 GHz radio imaging from the Nobeyama Radioheliograph (NoRH), spanning a little over two solar cycles (1992 - 2020). A tracer-independent method based on automated image correlation was employed on daily full-disc 17 GHz radio maps. Our results suggest that the upper chromosphere rotates significantly faster than the photosphere at all latitudes, with a relatively flatter latitudinal profile. A very weak anti-correlation between the equatorial rotation rate and solar activity is also observed. Our findings reaffirm the potential of radio observations to probe the dynamics of the solar chromosphere with reduced height ambiguity. The overlap of the equatorial rotation rate found in this study with that for $304$ {\AA} in the EUV regime lends additional support to the view that the equatorial rotation rates increase with height above the photosphere. Future coordinated studies at wavelengths with better-constrained height formation will be crucial for further understanding the complex dynamics of the solar atmosphere. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:57:42 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Routh",
"Srinjana",
""
],
[
"Kumari",
"Anshu",
""
],
[
"Pant",
"Vaibhav",
""
],
[
"Kandekar",
"Jaydeep",
""
],
[
"Banerjee",
"Dipankar",
""
],
[
"Khan",
"Mohd. Saleem",
""
],
[
"Mishra",
"Dibya Kirti",
""
]
] |
2507.02631 | Marc-Antoine Fardin | Marc-Antoine Fardin, Mathieu Hautefeuille, Vivek Sharma | Radical scaling: beyond our feet and fingers | 10 pages, 5 figures, 34 pages of supplementary information | null | null | null | cond-mat.soft physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Scaling laws arise and are eulogized across disciplines from natural to social sciences for providing pithy, quantitative, `scale-free', and `universal' power law relationships between two variables. On a log-log plot, the power laws display as straight lines, with a slope set by the exponent of the scaling law. In practice, a scaling relationship works only for a limited range, bookended by crossovers to other scaling laws. Leading with Taylor's oft-cited scaling law for the blast radius of an explosion against time, and by collating an unprecedented amount of datasets for laser-induced, chemical and nuclear explosions, we show distinct kinematics arise at the early and late stages. We illustrate that picking objective scales for the two axes using the transitions between regimes leads to the collapse of the data for the two regimes and their crossover, but the third regime is typically not mapped to the master curve. The objective scales permit us to abandon the arbitrarily chosen anthropocentric units of measurement, like feet for length and heart-beat for time, but the decimal system with ten digits (fingers) is still part of the picture. We show a remarkable collapse of all three regimes onto a common master curve occurs if we replace the base 10 by a dimensionless radix that combines the scales from the two crossovers. We also illustrate this approach of radical scaling for capillarity-driven pinching, coalescence and spreading of drops and bubbles, expecting such generalizations will be made for datasets across many disciplines. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 13:59:05 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fardin",
"Marc-Antoine",
""
],
[
"Hautefeuille",
"Mathieu",
""
],
[
"Sharma",
"Vivek",
""
]
] |
2507.02632 | ATLAS Collaboration | ATLAS Collaboration | Measurement of the top-quark pole mass in dileptonic $t\bar{t}+ 1\text{-jet}$ events at $\sqrt{s}=13$ TeV with the ATLAS experiment | 50 pages in total, author list starting page 33, 6 figures, 3 tables, submitted to JHEP. All figures including auxiliary figures are available at http://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/PAPERS/TOPQ-2018-23 | null | null | CERN-EP-2025-135 | hep-ex | http://creativecommons.org/licenses/by/4.0/ | A measurement of the top-quark pole mass $m_{t}^\text{pole}$ is presented in $t\bar{t}$ events with an additional jet, $t\bar{t}+1\text{-jet}$, produced in $pp$ collisions at $\sqrt{s}=13$ TeV. The data sample, recorded with the ATLAS experiment during Run 2 of the LHC, corresponds to an integrated luminosity of 140 $\text{fb}^{-1}$. Events with one electron and one muon of opposite electric charge in the final state are selected to measure the $t\bar{t}+1\text{-jet}$ differential cross-section as a function of the inverse of the invariant mass of the $t\bar{t}+1\text{-jet}$ system. Iterative Bayesian Unfolding is used to correct the data to enable comparison with fixed-order calculations at next-to-leading-order accuracy in the strong coupling. The process $pp \to t\bar{t}j$ ($2 \rightarrow 3$), where top quarks are taken as stable particles, and the process $pp \to b\bar{b}l^+\nu l^- \bar{\nu} j$ ($2 \to 7$), which includes top-quark decays to the dilepton final state and off-shell effects, are considered. The top-quark mass is extracted using a $\chi^2$ fit of the unfolded normalized differential cross-section distribution. The results obtained with the $2 \to 3$ and $2 \to 7$ calculations are compatible within theoretical uncertainties, providing an important consistency check. The more precise determination is obtained for the $2 \to 3 $ measurement: $m_{t}^\text{pole}=170.7\pm0.3(\text{stat.})\pm1.4(\text{syst.})\pm 0.3(\text{scale})\pm 0.2(\text{PDF}\oplus\alpha_\text{S})$ GeV, which is in good agreement with other top-quark mass results. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:00:00 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"ATLAS Collaboration",
"",
""
]
] |
2507.02633 | Martin Kroner | Le Liu, Igor Khanonkin, Johannes Eberle, Bernhard Rizek, Stefan F\"alt, Kenji Watanabe, Takashi Taniguchi, Ata\c{c} Imamo\u{g}lu, Martin Kroner | Single Photon Emitters in Ultra-Thin Hexagonal Boron Nitride Layers | null | null | null | null | cond-mat.mes-hall physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Single-photon emitters (SPE) in hexagonal boron nitride (h-BN) are promising for applications ranging from single-photon sources to quantum sensors. Previous studies exclusively focused on the generation and characterization of SPEs in relatively thick h-BN layers ($\geq$ 30 nm). However, for electrical and magnetic sensing applications, the thickness of the h-BN limits the attainable spatial resolution. Here, we report the observation of blue-wavelength emitters (B-centers) activated by electron beam irradiation in ultra-thin ($\simeq$ 3 nm) h-BN. These SPEs in ultra-thin flakes exhibit reduced brightness, broader zero-phonon line, and enhanced photobleaching. Remarkably, upon encapsulation in thicker h-BN, we restore their brightness, narrow linewidth 230$\mu$eV at 5K, resolution limited), suppress photobleaching, and confirm single-photon emission with $ g^{(2)}(0) < 0.4$ at room temperature. The possibility of generating SPEs in a few-layer h-BN and their subsequent incorporation into a van der Waals heterostructure paves the way for achieving quantum sensing with unprecedented nanometer-scale spatial resolution. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:00:50 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Le",
""
],
[
"Khanonkin",
"Igor",
""
],
[
"Eberle",
"Johannes",
""
],
[
"Rizek",
"Bernhard",
""
],
[
"Fält",
"Stefan",
""
],
[
"Watanabe",
"Kenji",
""
],
[
"Taniguchi",
"Takashi",
""
],
[
"Imamoğlu",
"Ataç",
""
],
[
"Kroner",
"Martin",
""
]
] |
2507.02634 | David Mguni | David H. Mguni | High-Order Deep Meta-Learning with Category-Theoretic Interpretation | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a new hierarchical deep learning framework for recursive higher-order meta-learning that enables neural networks (NNs) to construct, solve, and generalise across hierarchies of tasks. Central to this approach is a generative mechanism that creates \emph{virtual tasks} -- synthetic problem instances designed to enable the meta-learner to learn \emph{soft constraints} and unknown generalisable rules across related tasks. Crucially, this enables the framework to generate its own informative, task-grounded datasets thereby freeing machine learning (ML) training from the limitations of relying entirely on human-generated data. By actively exploring the virtual point landscape and seeking out tasks lower-level learners find difficult, the meta-learner iteratively refines constraint regions. This enhances inductive biases, regularises the adaptation process, and produces novel, unanticipated tasks and constraints required for generalisation. Each meta-level of the hierarchy corresponds to a progressively abstracted generalisation of problems solved at lower levels, enabling a structured and interpretable learning progression. By interpreting meta-learners as category-theoretic \emph{functors} that generate and condition a hierarchy of subordinate learners, we establish a compositional structure that supports abstraction and knowledge transfer across progressively generalised tasks. The category-theoretic perspective unifies existing meta-learning models and reveals how learning processes can be transformed and compared through functorial relationships, while offering practical design principles for structuring meta-learning. We speculate this architecture may underpin the next generation of NNs capable of autonomously generating novel, instructive tasks and their solutions, thereby advancing ML towards general artificial intelligence. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:01:14 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Mguni",
"David H.",
""
]
] |
2507.02635 | Yiwen Huang | Mao Luo, Zhi Wang, Yiwen Huang, Qingyun Zhang, Zhouxing Su, Zhipeng Lv, Wen Hu and Jianguo Li | SAT-BO: Verification Rule Learning and Optimization for FraudTransaction Detection | null | null | null | null | cs.CR cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Electronic payment platforms are estimated to process billions oftransactions daily, with the cumulative value of these transactionspotentially reaching into the trillions. Even a minor error within thishigh-volume environment could precipitate substantial financiallosses. To mitigate this risk, manually constructed verification rules,developed by domain experts, are typically employed to identifyand scrutinize transactions in production environments. However,due to the absence of a systematic approach to ensure the robust-ness of these verification rules against vulnerabilities, they remainsusceptible to exploitation.To mitigate this risk, manually constructed verification rules, de-veloped by domain experts, are typically employed to identify andscrutinize transactions in production environments. However, dueto the absence of a systematic approach to ensure the robustness ofthese verification rules against vulnerabilities, they remain suscep-tible to exploitation. To ensure data security, database maintainersusually compose complex verification rules to check whether aquery/update request is valid. However, the rules written by ex-perts are usually imperfect, and malicious requests may bypassthese rules. As a result, the demand for identifying the defects ofthe rules systematically emerges. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:01:26 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Luo",
"Mao",
""
],
[
"Wang",
"Zhi",
""
],
[
"Huang",
"Yiwen",
""
],
[
"Zhang",
"Qingyun",
""
],
[
"Su",
"Zhouxing",
""
],
[
"Lv",
"Zhipeng",
""
],
[
"Hu",
"Wen",
""
],
[
"Li",
"Jianguo",
""
]
] |
2507.02636 | Ning Qi | Ning Qi, Yousuf Baker, Bolun Xu | Online Convex Optimization for Coordinated Long-Term and Short-Term Isolated Microgrid Dispatch | null | null | null | null | math.OC cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | This paper proposes a novel non-anticipatory long-short-term coordinated dispatch framework for isolated microgrid with hybrid short-long-duration energy storages (LDES). We introduce a convex hull approximation model for nonconvex LDES electrochemical dynamics, facilitating computational tractability and accuracy. To address temporal coupling in SoC dynamics and long-term contracts, we generate hindsight-optimal state-of-charge (SoC) trajectories of LDES and netloads for offline training. In the online stage, we employ kernel regression to dynamically update the SoC reference and propose an adaptive online convex optimization (OCO) algorithm with SoC reference tracking and expert tracking to mitigate myopia and enable adaptive step-size optimization. We rigorously prove that both long-term and short-term policies achieve sublinear regret bounds over time, which improves with more regression scenarios, stronger tracking penalties, and finer convex approximations. Simulation results show that the proposed method outperforms state-of-the-art methods, reducing costs by 73.4%, eliminating load loss via reference tracking, and achieving an additional 2.4% cost saving via the OCO algorithm. These benefits scale up with longer LDES durations, and the method demonstrates resilience to poor forecasts and unexpected system faults. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:01:59 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Qi",
"Ning",
""
],
[
"Baker",
"Yousuf",
""
],
[
"Xu",
"Bolun",
""
]
] |
2507.02637 | Marco Piva | Marco Piva | Classification of $f(R)$ Theories Of Inflation And The Uniqueness of Starobinsky Model | 33 pages | null | null | null | hep-th astro-ph.CO gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We classify $f(R)$ theories using a mathematical analogy between slow-roll inflation and the renormalization-group flow. We derive the power spectra and spectral indices class by class and compare them with the latest data. The framework used for the classification allows us to determine the general structure of the $f(R)$ functions that belong to each class. Our main result is that only two classes survive. Moreover, we show that the Starobinsky model is the only polynomial $f(R)$ that can realize slow-roll inflation. In fact, all other polynomials belong to a special class that can only realize constant-roll inflation, at least far enough in the past. We point out some of the issues involved in considering a smooth transition between constant-roll and slow-roll inflation in this class of models. Finally, we derive the map that transforms the results from the Jordan frame to the Einstein frame. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:02:37 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Piva",
"Marco",
""
]
] |
2507.02638 | Mikhail Cherdantsev | Mikhail Cherdantsev, Andrey Piatnitski, Igor Velcic | Homogenisation and spectral convergence of high-contrast convolution type operators | 59 pages, 1 figure | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | The paper deals with homogenisation problems for high-contrast symmetric convolution-type operators with integrable kernels in media with a periodic microstructure. We adapt the two-scale convergence method to nonlocal convolution-type operators and obtain the homogenisation result both for problems stated in the whole space and in bounded domains with the homogeneous Dirichlet boundary condition.
Our main focus is on spectral analysis. We describe the spectrum of the limit two-scale operator and characterize the limit behaviour of the spectrum of the original problem as the microstructure period tends to zero. It is shown that the spectrum of the limit operator is a subset the limit of the spectrum of the original operator, and that they need not coincide. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:02:52 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Cherdantsev",
"Mikhail",
""
],
[
"Piatnitski",
"Andrey",
""
],
[
"Velcic",
"Igor",
""
]
] |
2507.02639 | Alberto Caron | Alberto Caron, Chris Hicks, Vasilios Mavroudis | On Efficient Bayesian Exploration in Model-Based Reinforcement Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we address the challenge of data-efficient exploration in reinforcement learning by examining existing principled, information-theoretic approaches to intrinsic motivation. Specifically, we focus on a class of exploration bonuses that targets epistemic uncertainty rather than the aleatoric noise inherent in the environment. We prove that these bonuses naturally signal epistemic information gains and converge to zero once the agent becomes sufficiently certain about the environment's dynamics and rewards, thereby aligning exploration with genuine knowledge gaps. Our analysis provides formal guarantees for IG-based approaches, which previously lacked theoretical grounding. To enable practical use, we also discuss tractable approximations via sparse variational Gaussian Processes, Deep Kernels and Deep Ensemble models. We then outline a general framework - Predictive Trajectory Sampling with Bayesian Exploration (PTS-BE) - which integrates model-based planning with information-theoretic bonuses to achieve sample-efficient deep exploration. We empirically demonstrate that PTS-BE substantially outperforms other baselines across a variety of environments characterized by sparse rewards and/or purely exploratory tasks. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:03:47 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Caron",
"Alberto",
""
],
[
"Hicks",
"Chris",
""
],
[
"Mavroudis",
"Vasilios",
""
]
] |
2507.02640 | Nina D\"ornemann | Nina D\"ornemann | Two-Sample Covariance Inference in High-Dimensional Elliptical Models | null | null | null | null | math.ST math.PR stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a two-sample test for large-dimensional covariance matrices in generalized elliptical models. The test statistic is based on a U-statistic estimator of the squared Frobenius norm of the difference between the two population covariance matrices. This statistic was originally introduced by Li and Chen (2012, AoS) for the independent component model. As a key theoretical contribution, we establish a new central limit theorem for the U-statistics under elliptical data, valid under both the null and alternative hypotheses. This result enables asymptotic control of the test level and facilitates a power analysis. To the best of our knowledge, the proposed test is the first such method to be supported by theoretical guarantees for elliptical data. Our approach imposes only mild assumptions on the covariance matrices and does neither require sparsity nor explicit growth conditions on the dimension-to-sample-size ratio. We illustrate our theoretical findings through applications to both synthetic and real-world data. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:07:03 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Dörnemann",
"Nina",
""
]
] |
2507.02641 | Shuaixin Yang | Shuaixin Yang, Yijia Li, Yue Xiao, Yong Liang Guan, Xianfu Lei, Zhiguo Ding | Pinching-Antenna-Assisted Index Modulation: Channel Modeling, Transceiver Design, and Performance Analysis | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, a novel pinching-antenna assisted index modulation (PA-IM) scheme is proposed for improving the spectral efficiency without increasing the hardware complexity, where the information bits are conveyed not only by the conventional M-ary quadrature amplitude modulation (QAM) symbols but also by the indices of pinching antenna (PA) position patterns. To realize the full potential of this scheme, this paper focuses on the comprehensive transceiver design, addressing key challenges in signal detection at the receiver and performance optimization at thetransmitter. First, a comprehensive channel model is formulated for this architecture, which sophisticatedly integrates the deterministic in-waveguide propagation effects with the stochastic nature of wireless channels, including both largescale path loss and small-scale fading. Next, to overcome the prohibitive complexity of optimal maximum likelihood (ML) detection, a low-complexity box-optimized sphere decoding (BOSD) algorithm is designed, which adaptively prunes the search space whilst preserving optimal ML performance. Furthermore, an analytical upper bound on the bit error rate (BER) is derived and validated by the simulations. Moreover, a new transmit precoding method is designed using manifold optimization, which minimizes the BER by jointly optimizing the complex-valued precoding coefficients across the waveguides for the sake of maximizing the minimum Euclidean distance of all received signal points. Finally, the simulation results demonstrate that the proposed PA-IM scheme attains a significant performance gain over its conventional counterparts and that the overall BER of the pinching-antenna system is substantially improved by the proposed precoding design. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:07:45 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yang",
"Shuaixin",
""
],
[
"Li",
"Yijia",
""
],
[
"Xiao",
"Yue",
""
],
[
"Guan",
"Yong Liang",
""
],
[
"Lei",
"Xianfu",
""
],
[
"Ding",
"Zhiguo",
""
]
] |
2507.02642 | Tomas Polcar | Jan Komeda, Antonio Cammarata, Tomas Polcar | Optimal boron-doped graphene substrate for glucose Raman signal enhancement | null | null | null | null | cond-mat.mtrl-sci physics.app-ph physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | Surface Enhanced Raman Spectroscopy (SERS) is a highly sensitive and selective technique that greatly enhances the signal of an analyte, compared with its signal from classical Raman Spectroscopy, due to its interaction with a substrates surface. It has been shown that low concentration boron-doped graphene (B-graphene) enhances the Raman signal of simple organic molecules like pyridine. Recent studies also suggest that B-graphene can remain thermodynamically stable when doped with significantly higher concentrations of boron than previously observed. In this framework, we use quantum mechanical simulations to investigate the influence of dopant concentration and geometric distribution on the effectiveness of B-doped graphene as a SERS substrate, with glucose as analyte. By combining analysis of interatomic force constants and of phonon eigenvectors composition, we conclude that higher doping concentrations provide a larger enhancement to glucose's Raman signal, while the molecule orientation relative to the surface plays a fundamental role in the Raman response. We suggest that high concentration B-graphene presents itself as a potential substrate for SERS based detection of glucose, while the used phonon-based analysis can be promptly applied for the search of promising candidates as substrate materials for enhanced Raman response. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:07:47 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Komeda",
"Jan",
""
],
[
"Cammarata",
"Antonio",
""
],
[
"Polcar",
"Tomas",
""
]
] |
2507.02643 | Diego Corr\^ea Da Silva | Diego Corr\^ea da Silva and Dietmar Jannach | Calibrated Recommendations: Survey and Future Directions | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | The idea of calibrated recommendations is that the properties of the items that are suggested to users should match the distribution of their individual past preferences. Calibration techniques are therefore helpful to ensure that the recommendations provided to a user are not limited to a certain subset of the user's interests. Over the past few years, we have observed an increasing number of research works that use calibration for different purposes, including questions of diversity, biases, and fairness. In this work, we provide a survey on the recent developments in the area of calibrated recommendations. We both review existing technical approaches for calibration and provide an overview on empirical and analytical studies on the effectiveness of calibration for different use cases. Furthermore, we discuss limitations and common challenges when implementing calibration in practice. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:08:10 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"da Silva",
"Diego Corrêa",
""
],
[
"Jannach",
"Dietmar",
""
]
] |
2507.02644 | Yuntian Gu | Yuntian Gu, Wenrui Li, Heng Lin, Bo Zhan, Ruichen Li, Yifei Huang, Di He, Yantao Wu, Tao Xiang, Mingpu Qin, Liwei Wang, Dingshun Lv | Solving the Hubbard model with Neural Quantum States | null | null | null | null | cond-mat.str-el cs.AI quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid development of neural quantum states (NQS) has established it as a promising framework for studying quantum many-body systems. In this work, by leveraging the cutting-edge transformer-based architectures and developing highly efficient optimization algorithms, we achieve the state-of-the-art results for the doped two-dimensional (2D) Hubbard model, arguably the minimum model for high-Tc superconductivity. Interestingly, we find different attention heads in the NQS ansatz can directly encode correlations at different scales, making it capable of capturing long-range correlations and entanglements in strongly correlated systems. With these advances, we establish the half-filled stripe in the ground state of 2D Hubbard model with the next nearest neighboring hoppings, consistent with experimental observations in cuprates. Our work establishes NQS as a powerful tool for solving challenging many-fermions systems. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:08:25 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Gu",
"Yuntian",
""
],
[
"Li",
"Wenrui",
""
],
[
"Lin",
"Heng",
""
],
[
"Zhan",
"Bo",
""
],
[
"Li",
"Ruichen",
""
],
[
"Huang",
"Yifei",
""
],
[
"He",
"Di",
""
],
[
"Wu",
"Yantao",
""
],
[
"Xiang",
"Tao",
""
],
[
"Qin",
"Mingpu",
""
],
[
"Wang",
"Liwei",
""
],
[
"Lv",
"Dingshun",
""
]
] |
2507.02645 | Harry Cheng | Harry Cheng and Ming-Hui Liu and Yangyang Guo and Tianyi Wang and Liqiang Nie and Mohan Kankanhalli | Fair Deepfake Detectors Can Generalize | 14 pages, version 1 | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deepfake detection models face two critical challenges: generalization to unseen manipulations and demographic fairness among population groups. However, existing approaches often demonstrate that these two objectives are inherently conflicting, revealing a trade-off between them. In this paper, we, for the first time, uncover and formally define a causal relationship between fairness and generalization. Building on the back-door adjustment, we show that controlling for confounders (data distribution and model capacity) enables improved generalization via fairness interventions. Motivated by this insight, we propose Demographic Attribute-insensitive Intervention Detection (DAID), a plug-and-play framework composed of: i) Demographic-aware data rebalancing, which employs inverse-propensity weighting and subgroup-wise feature normalization to neutralize distributional biases; and ii) Demographic-agnostic feature aggregation, which uses a novel alignment loss to suppress sensitive-attribute signals. Across three cross-domain benchmarks, DAID consistently achieves superior performance in both fairness and generalization compared to several state-of-the-art detectors, validating both its theoretical foundation and practical effectiveness. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:10:02 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Cheng",
"Harry",
""
],
[
"Liu",
"Ming-Hui",
""
],
[
"Guo",
"Yangyang",
""
],
[
"Wang",
"Tianyi",
""
],
[
"Nie",
"Liqiang",
""
],
[
"Kankanhalli",
"Mohan",
""
]
] |
2507.02646 | Jaewon Chang | Jaewon Chang | Distinguished bases in the wrapped Floer cohomology of tropical Lagrangian surfaces | null | null | null | null | math.SG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a new construction for tropical Lagrangian surfaces in $(\mathbb{C}^*)^2$. This construction makes the surfaces special Lagrangian, which gives a strong control over the asymptotic behavior of holomorphic disks near each cylindrical end. As a result, we find a distinguished basis of the wrapped Floer cohomology ring $HW^\bullet(L_f, L_f)$ for several different examples and find the corresponding distinguished bases of functions on the mirror curves. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:11:48 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Chang",
"Jaewon",
""
]
] |
2507.02647 | Nathan Hagen | Luis Garza-Soto and Nathan Hagen | The geometric phase of rotations and 3D coordinate transformations | null | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | The wave description of geometric phase uses the superposition of light waves to explain the geometric phase's origin. While our previous work focused on a basis of linearly polarized waves, here we show that the same concepts can be applied to circularly polarized waves, and to any case in which a rotator is itself subjected to rotation. As with a linear polarization basis, we show that the addition of two vectors (rotators) with different orientations and magnitudes causes the orientation of the resulting vector to shift towards the component vector of greater magnitude, i.e. it introduces a geometric phase. We illustrate this approach with two classic examples of the geometric phase of rotations in space: a system of three fold mirrors, and the helical coiled fiber. In both cases we show that it is possible to derive the phase shift directly from the electromagnetic wave vector without needing to resort to mathematical abstractions such as differential geometry, or calculating solid angles in the space of directions. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:12:35 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Garza-Soto",
"Luis",
""
],
[
"Hagen",
"Nathan",
""
]
] |
2507.02648 | Aspen Hopkins | Aspen K. Hopkins, Isabella Struckman, Kevin Klyman, Susan S. Silbey | Recourse, Repair, Reparation, & Prevention: A Stakeholder Analysis of AI Supply Chains | null | null | 10.1145/3715275.3732017. | null | cs.CY | http://creativecommons.org/licenses/by/4.0/ | The AI industry is exploding in popularity, with increasing attention to potential harms and unwanted consequences. In the current digital ecosystem, AI deployments are often the product of AI supply chains (AISC): networks of outsourced models, data, and tooling through which multiple entities contribute to AI development and distribution. AI supply chains lack the modularity, redundancies, or conventional supply chain practices that enable identification, isolation, and easy correction of failures, exacerbating the already difficult processes of responding to ML-generated harms. As the stakeholders participating in and impacted by AISCs have scaled and diversified, so too have the risks they face. In this stakeholder analysis of AI supply chains, we consider who participates in AISCs, what harms they face, where sources of harm lie, and how market dynamics and power differentials inform the type and probability of remedies. Because AI supply chains are purposely invented and implemented, they may be designed to account for, rather than ignore, the complexities, consequences, and risks of deploying AI systems. To enable responsible design and management of AISCs, we offer a typology of responses to AISC-induced harms: recourse, repair, reparation or prevention. We apply this typology to stakeholders participating in a health-care AISC across three stylized markets $\unicode{x2013}$ vertical integration, horizontal integration, free market $\unicode{x2013}$ to illustrate how stakeholder positioning and power within an AISC may shape responses to an experienced harm. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:12:43 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Hopkins",
"Aspen K.",
""
],
[
"Struckman",
"Isabella",
""
],
[
"Klyman",
"Kevin",
""
],
[
"Silbey",
"Susan S.",
""
]
] |
2507.02649 | Sunder Ram Krishnan | Sunder Ram Krishnan | Restricted Quasiconvexity Isometry Property for Symmetric $\alpha$-Stable Random Matrices | 10 pages | null | null | null | math.PR eess.SP | http://creativecommons.org/licenses/by/4.0/ | We formulate a generalization of the Restricted Isometry Property (RIP) referred to as the Restricted Quasiconvexity Isometry Property (RQIP) for alpha stable random projections with $0<\alpha<1$. A lower bound on the number of rows for RQIP to hold for random matrices whose entries are drawn from a symmetric $\alpha$-stable ($S\alpha S$) distribution is derived. The proof leverages two key components: a concentration inequality for empirical fractional moments of $S\alpha S$ variables and a covering number bound for sparse $\ell_\alpha$ balls. The resulting sample complexity reflects the polynomial tail behavior of the concentration and reinforces an observation made in the literature that the RIP framework may have to be replaced with other sparse recovery formulations in practice, such as those based on the null space property. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:16:34 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Krishnan",
"Sunder Ram",
""
]
] |
2507.02650 | Lizhu Sun | Jueru Liu, Lizhu Sun, Changjiang Bu | The ordering of hypertrees and unicyclic hypergraphs by the traces of $\mathcal{A}_{\alpha}$-tensor | null | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | For a real number $\alpha\in[0,1]$ and a $k$-uniform hypergraph $\mathcal{H}$, $\mathcal{A}_{\alpha}(\mathcal{H})=\alpha\mathcal{D}(\mathcal{H})+(1-\alpha)\mathcal{A}(\mathcal{H})$ is called the $\mathcal{A}_{\alpha}$-tensor of $\mathcal{H}$, where $\mathcal{D}(\mathcal{H})$ and $\mathcal{A}(\mathcal{H})$ are the degree tensor and adjacency tensor of $\mathcal{H}$, respectively. The sum of the $d$-th powers of all eigenvalues of $\mathcal{A}_{\alpha}(\mathcal{H})$ is called the $d$-th order $\mathcal{A}_{\alpha}$-spectral moment of $\mathcal{H}$, which is equal to the $d$-th order trace of $\mathcal{A}_{\alpha}(\mathcal{H})$. In this paper, some hypergraphs are ordered lexicographically by their $\mathcal{A}_{\alpha}$-spectral moments in non-decreasing order. The first, the second, the last and the second last hypergraphs among all $k$-uniform linear unicyclic hypergraphs and hypertrees are characterized, respectively. We give the first and the last hypergraphs among all $k$-uniform linear unicyclic hypergraphs with given grith, and characterize the last hypertree among all $k$-uniform hypertrees with given diameter. Furthermore, we determine some extreme values of the $\mathcal{A}_{\alpha}$-spectral moments for hypertrees and linear unicyclic hypergraphs, respectively. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:16:57 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Jueru",
""
],
[
"Sun",
"Lizhu",
""
],
[
"Bu",
"Changjiang",
""
]
] |
2507.02651 | Xin Zhang | Meng-Lin Zhao, Yue Shao, Sai Wang, Xin Zhang | Prospects for probing dark matter particles and primordial black holes with the Square Kilometre Array using the 21 cm power spectrum at cosmic dawn | 26 pages, 11 figures | null | null | null | astro-ph.CO gr-qc hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Probing the nature of dark matter (DM) remains an outstanding problem in modern cosmology. The 21 cm signal, as a sensitive tracer of neutral hydrogen during cosmic dawn, provides a unique means to investigate DM nature during this critical epoch. Annihilation and decay of DM particles, as well as Hawking radiation of primordial black holes (PBHs), can modify the thermal and ionization histories of the early universe, leaving distinctive imprints on the 21 cm power spectrum. Therefore, the redshifted 21 cm power spectrum serves as a powerful tool to investigate such DM processes. In this work, we systematically assess the potential of the upcoming Square Kilometre Array (SKA) to constrain DM and PBH parameters using the 21 cm power spectrum. Assuming $10,000$ hours of integration time, the SKA is projected to reach sensitivities of $\langle\sigma v\rangle \leq 10^{-28}\,{\rm cm}^{3}\,{\rm s}^{-1}$ and $\tau\geq 10^{28}\,{\rm seconds}$, for $10\,{\rm GeV}$ DM particles. It can also probe PBHs with masses of $10^{16}\,\mathrm{g}$ and abundances $f_{\mathrm{PBH}} \leq 10^{-6}$. These results indicate that the SKA could place constraints on DM annihilation, decay, and PBH Hawking radiation that are up to two to three orders of magnitude stronger than current limits. Furthermore, the SKA is expected to exceed existing bounds on sub-GeV DM and to probe Hawking radiation from PBHs with masses above $10^{17}\,{\rm g}$, which are otherwise inaccessible by conventional cosmological probes. Overall, the SKA holds great promise for advancing our understanding of both DM particles and PBHs, potentially offering new insights into the fundamental nature of DM. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:17:13 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhao",
"Meng-Lin",
""
],
[
"Shao",
"Yue",
""
],
[
"Wang",
"Sai",
""
],
[
"Zhang",
"Xin",
""
]
] |
2507.02652 | Jiajie Jin | Jiajie Jin, Xiaoxi Li, Guanting Dong, Yuyao Zhang, Yutao Zhu, Yang Zhao, Hongjin Qian, Zhicheng Dou | Decoupled Planning and Execution: A Hierarchical Reasoning Framework for Deep Search | 9 pages | null | null | null | cs.AI cs.CL cs.IR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Complex information needs in real-world search scenarios demand deep reasoning and knowledge synthesis across diverse sources, which traditional retrieval-augmented generation (RAG) pipelines struggle to address effectively. Current reasoning-based approaches suffer from a fundamental limitation: they use a single model to handle both high-level planning and detailed execution, leading to inefficient reasoning and limited scalability. In this paper, we introduce HiRA, a hierarchical framework that separates strategic planning from specialized execution. Our approach decomposes complex search tasks into focused subtasks, assigns each subtask to domain-specific agents equipped with external tools and reasoning capabilities, and coordinates the results through a structured integration mechanism. This separation prevents execution details from disrupting high-level reasoning while enabling the system to leverage specialized expertise for different types of information processing. Experiments on four complex, cross-modal deep search benchmarks demonstrate that HiRA significantly outperforms state-of-the-art RAG and agent-based systems. Our results show improvements in both answer quality and system efficiency, highlighting the effectiveness of decoupled planning and execution for multi-step information seeking tasks. Our code is available at https://github.com/ignorejjj/HiRA. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:18:08 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Jin",
"Jiajie",
""
],
[
"Li",
"Xiaoxi",
""
],
[
"Dong",
"Guanting",
""
],
[
"Zhang",
"Yuyao",
""
],
[
"Zhu",
"Yutao",
""
],
[
"Zhao",
"Yang",
""
],
[
"Qian",
"Hongjin",
""
],
[
"Dou",
"Zhicheng",
""
]
] |
2507.02653 | Andraz Omahen | Andraz Omahen, Simon Storz, Marius Bild, Dario Scheiwiller, Matteo Fadel, Yiwen Chu | An Ultra-Cold Mechanical Quantum Sensor for Tests of New Physics | null | null | null | null | quant-ph cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | Initialization of mechanical modes in the quantum ground state is crucial for their use in quantum information and quantum sensing protocols. In quantum processors, impurity of the modes' initial state affects the infidelity of subsequent quantum algorithms. In quantum sensors, excitations out of the ground state contribute to the noise of the detector, and their prevalence puts a bound on rare events that deposit energy into the mechanical modes. In this work, we measure the excited-state populations of GHz-frequency modes in a high-overtone bulk acoustic wave resonator (HBAR). We find that the population of the first excited state can be as low as $P_p$=(1.2$\pm$5.5)$\times10^{-5}$, corresponding to an effective temperature of 25.2 mK, which are upper bounds limited by imperfections in the measurement process. These results compare favorably to the lowest populations measured in superconducting circuits. Finally, we use the measured populations to constrain the amplitude of high-frequency gravitational waves, the kinetic mixing strength of ultra-light dark matter, and non-linear modifications of the Schr\"{o}dinger equation describing wavefunction collapse mechanisms. Our work establishes HBARs as a versatile resource for quantum state initialization and studies of fundamental physics. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:18:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Omahen",
"Andraz",
""
],
[
"Storz",
"Simon",
""
],
[
"Bild",
"Marius",
""
],
[
"Scheiwiller",
"Dario",
""
],
[
"Fadel",
"Matteo",
""
],
[
"Chu",
"Yiwen",
""
]
] |
2507.02654 | Rui Xie | Rui Xie, Asad Ul Haq, Yunhua Fang, Linsen Ma, Sanchari Sen, Swagath Venkataramani, Liu Liu, Tong Zhang | Breaking the HBM Bit Cost Barrier: Domain-Specific ECC for AI Inference Infrastructure | null | null | null | null | cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-Bandwidth Memory (HBM) delivers exceptional bandwidth and energy efficiency for AI workloads, but its high cost per bit, driven in part by stringent on-die reliability requirements, poses a growing barrier to scalable deployment. This work explores a system-level approach to cost reduction by eliminating on-die ECC and shifting all fault management to the memory controller. We introduce a domain-specific ECC framework combining large-codeword Reed--Solomon~(RS) correction with lightweight fine-grained CRC detection, differential parity updates to mitigate write amplification, and tunable protection based on data importance. Our evaluation using LLM inference workloads shows that, even under raw HBM bit error rates up to $10^{-3}$, the system retains over 78\% of throughput and 97\% of model accuracy compared with systems equipped with ideal error-free HBM. By treating reliability as a tunable system parameter rather than a fixed hardware constraint, our design opens a new path toward low-cost, high-performance HBM deployment in AI infrastructure. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:18:41 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Xie",
"Rui",
""
],
[
"Haq",
"Asad Ul",
""
],
[
"Fang",
"Yunhua",
""
],
[
"Ma",
"Linsen",
""
],
[
"Sen",
"Sanchari",
""
],
[
"Venkataramani",
"Swagath",
""
],
[
"Liu",
"Liu",
""
],
[
"Zhang",
"Tong",
""
]
] |
2507.02655 | Mario Bebendorf | S. Aziz and M. Bauer and M. Bebendorf and T. Rau | On low-dimensional approximation of function spaces of interior regularity | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Many elliptic boundary value problems exhibit an interior regularity property, which can be exploited to construct local approximation spaces that converge exponentially within function spaces satisfying this property. These spaces can be used to define local ansatz spaces within the framework of generalised finite element methods, leading to a better relation between dimensionality and convergence order. In this paper, we present a new technique for the construction of such spaces for Lipschitz domains. Instead of the commonly used approach based on eigenvalue problems it relies on extensions of approximations performed on the boundary. Hence, it improves the influence of the spatial dimension on the exponential convergence and allows to construct the local spaces by solving the original kind of variational problems on easily structured domains. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:18:48 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Aziz",
"S.",
""
],
[
"Bauer",
"M.",
""
],
[
"Bebendorf",
"M.",
""
],
[
"Rau",
"T.",
""
]
] |
2507.02656 | Harrison Nicholls | Harrison Nicholls, Tim Lichtenberg, Richard D. Chatterjee, Claire Marie Guimond, Emma Postolec, Raymond T. Pierrehumbert | Volatile-rich evolution of molten super-Earth L 98-59 d | In review. 4 figures and 4 pages in the main text | null | null | null | astro-ph.EP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Small low-density exoplanets are sculpted by strong stellar irradiation but their origin is unknown. Two competing scenarios are that they formed either with rocky interiors and H2-He atmospheres (`gas-dwarfs'), or instead with bulk compositions dominated by H2O phases (`water-worlds'). Observations of L 98-59 d have revealed its unusually low density and spectral characteristics, opening a window for disentangling its origin. We constrain the possible range of evolutionary histories linking L 98-59 d's birth conditions to these observations, using a coupled atmosphere-interior modelling framework. Observations are explained by a chemically-reducing mantle and early substantial (>1.8 mass%) sulfur and hydrogen content, inconsistent with both the gas-dwarf and water-world origin scenarios. Observed spectral features are explained by in-situ photochemical production of SO2 in an H2 background. L 98-59 d's interior comprises a permanent magma ocean, allowing long-term retention of volatiles within its interior over billions of years, consistent with surveyed Kepler Survey trends. Modelling of the `cosmic shoreline' suggests that early volatile-metal budget is a determinant of atmospheric retention - here, we explicitly show that large initial budgets can explain JWST observations of some super-Earth planets (including DDT targets). In breaking with the supposed gas-dwarf/water-world dichotomy, we reveal an evolutionary pathway defined by high-molar-mass atmospheres overlying magma oceans, and invite a more nuanced taxonomy of small exoplanets. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:19:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Nicholls",
"Harrison",
""
],
[
"Lichtenberg",
"Tim",
""
],
[
"Chatterjee",
"Richard D.",
""
],
[
"Guimond",
"Claire Marie",
""
],
[
"Postolec",
"Emma",
""
],
[
"Pierrehumbert",
"Raymond T.",
""
]
] |
2507.02657 | Jens Schl\"oter | Jens Schl\"oter | On the Complexity of Knapsack under Explorable Uncertainty: Hardness and Algorithms | null | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the knapsack problem under explorable uncertainty, we are given a knapsack instance with uncertain item profits. Instead of having access to the precise profits, we are only given uncertainty intervals that are guaranteed to contain the corresponding profits. The actual item profit can be obtained via a query. The goal of the problem is to adaptively query item profits until the revealed information suffices to compute an optimal (or approximate) solution to the underlying knapsack instance. Since queries are costly, the objective is to minimize the number of queries.
In the offline variant of this problem, we assume knowledge of the precise profits and the task is to compute a query set of minimum cardinality that a third party without access to the profits could use to identify an optimal (or approximate) knapsack solution. We show that this offline variant is complete for the second-level of the polynomial hierarchy, i.e., $\Sigma_2^p$-complete, and cannot be approximated within a non-trivial factor unless $\Sigma_2^p = \Delta_2^p$. Motivated by these strong hardness results, we consider a resource-augmented variant of the problem where the requirements on the query set computed by an algorithm are less strict than the requirements on the optimal solution we compare against. More precisely, a query set computed by the algorithm must reveal sufficient information to identify an approximate knapsack solution, while the optimal query set we compare against has to reveal sufficient information to identify an optimal solution. We show that this resource-augmented setting allows interesting non-trivial algorithmic results. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:19:41 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Schlöter",
"Jens",
""
]
] |
2507.02658 | Chuliang Fu | Thanh Nguyen, Chuliang Fu, Mouyang Cheng, Buxuan Li, Tyra E. Espedal, Zhantao Chen, Kuan Qiao, Kumar Neeraj, Abhijatmedhi Chotrattanapituk, Denisse Cordova Carrizales, Eunbi Rha, Tongtong Liu, Shivam N. Kajale, Deblina Sarkar, Donald A. Walko, Haidan Wen, Svetlana V. Boriskina, Gang Chen, Jeehwan Kim, and Mingda Li | Spatiotemporal Mapping of Anisotropic Thermal Transport in GaN Thin Films via Ultrafast X-ray Diffraction | 24 pages, 4 figures | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Efficient thermal management is essential for the reliability of modern power electronics, where increasing device density leads to severe heat dissipation challenges. However, in thin-film systems, thermal transport is often compromised by interfacial resistance and microscale defects introduced during synthesis or transfer, which are difficult to characterize using conventional techniques. Here we present a non-contact, spatiotemporal-resolved ultrafast x-ray diffraction method to extract in-plane thermal conductivity and thermal boundary conductance, using GaN thin films on silicon as a model system. By tracking the pump-induced lattice strain, we reconstruct the lateral heat flow dynamics and quantitatively probe thermal transport near a wrinkle defect. We uncover pronounced asymmetric heat dissipation across the wrinkle, with a four-fold reduction in the local thermal conductivity near the wrinkle and a 25% drop in interfacial conductance. Our work demonstrates that ultrafast x-ray diffraction can serve as a precise thermal metrology tool for characterizing heat transport in multilayered thin-film structures for next-generation microelectronic devices. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:20:21 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Nguyen",
"Thanh",
""
],
[
"Fu",
"Chuliang",
""
],
[
"Cheng",
"Mouyang",
""
],
[
"Li",
"Buxuan",
""
],
[
"Espedal",
"Tyra E.",
""
],
[
"Chen",
"Zhantao",
""
],
[
"Qiao",
"Kuan",
""
],
[
"Neeraj",
"Kumar",
""
],
[
"Chotrattanapituk",
"Abhijatmedhi",
""
],
[
"Carrizales",
"Denisse Cordova",
""
],
[
"Rha",
"Eunbi",
""
],
[
"Liu",
"Tongtong",
""
],
[
"Kajale",
"Shivam N.",
""
],
[
"Sarkar",
"Deblina",
""
],
[
"Walko",
"Donald A.",
""
],
[
"Wen",
"Haidan",
""
],
[
"Boriskina",
"Svetlana V.",
""
],
[
"Chen",
"Gang",
""
],
[
"Kim",
"Jeehwan",
""
],
[
"Li",
"Mingda",
""
]
] |
2507.02659 | Shaojie Zhuo | Ramchalam Kinattinkara Ramakrishnan, Zhaocong Yuan, Shaojie Zhuo, Chen Feng, Yicheng Lin, Chenzheng Su, Xiaopeng Zhang | OmniDraft: A Cross-vocabulary, Online Adaptive Drafter for On-device Speculative Decoding | null | null | null | null | cs.LG cs.CL | http://creativecommons.org/licenses/by/4.0/ | Speculative decoding generally dictates having a small, efficient draft model that is either pretrained or distilled offline to a particular target model series, for instance, Llama or Qwen models. However, within online deployment settings, there are two major challenges: 1) usage of a target model that is incompatible with the draft model; 2) expectation of latency improvements over usage and time. In this work, we propose OmniDraft, a unified framework that enables a single draft model to operate with any target model and adapt dynamically to user data. We introduce an online n-gram cache with hybrid distillation fine-tuning to address the cross-vocabulary mismatch across draft and target models; and further improve decoding speed by leveraging adaptive drafting techniques. OmniDraft is particularly suitable for on-device LLM applications where model cost, efficiency and user customization are the major points of contention. This further highlights the need to tackle the above challenges and motivates the \textit{``one drafter for all''} paradigm. We showcase the proficiency of the OmniDraft framework by performing online learning on math reasoning, coding and text generation tasks. Notably, OmniDraft enables a single Llama-68M model to pair with various target models including Vicuna-7B, Qwen2-7B and Llama3-8B models for speculative decoding; and additionally provides up to 1.5-2x speedup. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:20:41 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ramakrishnan",
"Ramchalam Kinattinkara",
""
],
[
"Yuan",
"Zhaocong",
""
],
[
"Zhuo",
"Shaojie",
""
],
[
"Feng",
"Chen",
""
],
[
"Lin",
"Yicheng",
""
],
[
"Su",
"Chenzheng",
""
],
[
"Zhang",
"Xiaopeng",
""
]
] |
2507.02660 | Aman Kumar | Deepak Narayan Gadde, Keerthan Kopparam Radhakrishna, Vaisakh Naduvodi Viswambharan, Aman Kumar, Djones Lettnin, Wolfgang Kunz, Sebastian Simon | Hey AI, Generate Me a Hardware Code! Agentic AI-based Hardware Design & Verification | To appear at the 38th SBC/SBMicro/IEEE Symposium on Integrated Circuits and Systems Design (SBCCI), August 25-29, 2025, Manaus, BRAZIL | null | null | null | cs.AI cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern Integrated Circuits (ICs) are becoming increasingly complex, and so is their development process. Hardware design verification entails a methodical and disciplined approach to the planning, development, execution, and sign-off of functionally correct hardware designs. This tedious process requires significant effort and time to ensure a bug-free tape-out. The field of Natural Language Processing has undergone a significant transformation with the advent of Large Language Models (LLMs). These powerful models, often referred to as Generative AI (GenAI), have revolutionized how machines understand and generate human language, enabling unprecedented advancements in a wide array of applications, including hardware design verification. This paper presents an agentic AI-based approach to hardware design verification, which empowers AI agents, in collaboration with Humain-in-the-Loop (HITL) intervention, to engage in a more dynamic, iterative, and self-reflective process, ultimately performing end-to-end hardware design and verification. This methodology is evaluated on five open-source designs, achieving over 95% coverage with reduced verification time while demonstrating superior performance, adaptability, and configurability. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:20:57 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Gadde",
"Deepak Narayan",
""
],
[
"Radhakrishna",
"Keerthan Kopparam",
""
],
[
"Viswambharan",
"Vaisakh Naduvodi",
""
],
[
"Kumar",
"Aman",
""
],
[
"Lettnin",
"Djones",
""
],
[
"Kunz",
"Wolfgang",
""
],
[
"Simon",
"Sebastian",
""
]
] |
2507.02661 | Signe Lundqvist | Daniel Irving Bernstein and Signe Lundqvist | The pure condition for incidence geometries | 14 pages, 4 figures | null | null | null | math.CO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The space of \emph{parallel redrawings} of an incidence geometry $(P,H,I)$ with an assigned set of normals is the set of points and hyperplanes in $\mathbb{R}^d$ satisfying the incidences given by $(P,H,I)$, such that the hyperplanes have the assigned normals.
In 1989, Whiteley characterized the incidence geometries that have d-dimensional realizations with generic hyperplane normals such that all points and hyperplanes are distinct. However, some incidence geometries can be realized as points and hyperplanes in d-dimensional space, with the points and hyperplanes distinct, but only for specific choices of normals. Such incidence geometries are the topic of this article.
In this article, we introduce a pure condition for parallel redrawings of incidence geometries, analogous to the pure condition for bar-and-joint frameworks, introduced by White and Whiteley. The d-dimensional pure condition of an incidence geometry (P,H,I) imposes a condition on the normals assigned to the hyperplanes of (P,H,I) required for d-dimensional realizations of (P,H,I) with distinct points. We use invariant theory to show that is a bracket polynomial. We will also explicitly compute the pure condition as a bracket polynomial for some examples in the plane. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:21:52 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Bernstein",
"Daniel Irving",
""
],
[
"Lundqvist",
"Signe",
""
]
] |
2507.02662 | Antoine Mouzard | Quentin Chauleur, Antoine Mouzard | High order uniform in time schemes for weakly nonlinear Schr\"odinger equation and wave turbulence | null | null | null | null | math.NA cs.NA math.AP | http://creativecommons.org/licenses/by/4.0/ | We introduce two multiscale numerical schemes for the time integration of weakly nonlinear Schr\"odinger equations, built upon the discretization of Picard iterates of the solution. These high-order schemes are designed to achieve high precision with respect to the small nonlinearity parameter under particular CFL condition. By exploiting the scattering properties of these schemes thanks to a low-frequency projected linear flow, we also establish its uniform accuracy over long time horizons. Numerical simulations are provided to illustrate the theoretical results, and these schemes are further applied to investigate dynamics in the framework of wave turbulence. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:23:15 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Chauleur",
"Quentin",
""
],
[
"Mouzard",
"Antoine",
""
]
] |
2507.02663 | Yongjiang Liu | Yongjiang Liu, Haoxi Li, Xiaosong Ma, Jie Zhang, Song Guo | Think How to Think: Mitigating Overthinking with Autonomous Difficulty Cognition in Large Reasoning Models | 21 pages, 18 figures | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent Long Reasoning Models(LRMs) have demonstrated remarkable capabilities in handling complex reasoning tasks, but are hindered by excessive overthinking. To explore its essence, our empirical analysis reveals that LRMs are primarily limited to recognizing task properties (i.e., difficulty levels) like humans before solving the problem, leading to a one-size-fits-all reasoning process. Inspired by this, a pressing and natural question emerges: Can we bootstrap such ability to further alleviate the overthinking phenomenon in LRMs? In this paper, we propose Think-How-to-Think (TH2T), a novel two-stage fine-tuning strategy that progressively inspires LRMs' difficulty cognition and redundancy cognition. First, we introduce difficulty-hypnosis in the prefixes of model outputs to intervene in the internal reasoning trajectory. Combined with a heterogeneous short and long reasoning dataset, the trained model enhances its sensitivity to task difficulty, enabling native, differentiated reasoning strategies across various tasks. Second, we further extend redundancy-hypnosis to the internal reasoning process, guiding the model to identify redundant structures within the reasoning steps and generate more concise reasoning outputs. Experiments on 7B/14B/32B models demonstrate that TH2T significantly reduces inference costs (more than 70% on easy tasks and 40% on hard tasks) while maintaining performance stability. The resulting outputs exhibit clear difficulty-aware capabilities and reduced redundancy (e.g., reflection). | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:24:26 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Yongjiang",
""
],
[
"Li",
"Haoxi",
""
],
[
"Ma",
"Xiaosong",
""
],
[
"Zhang",
"Jie",
""
],
[
"Guo",
"Song",
""
]
] |
2507.02664 | Ziyin Zhou | Ziyin Zhou, Yunpeng Luo, Yuanchen Wu, Ke Sun, Jiayi Ji, Ke Yan, Shouhong Ding, Xiaoshuai Sun, Yunsheng Wu, Rongrong Ji | AIGI-Holmes: Towards Explainable and Generalizable AI-Generated Image Detection via Multimodal Large Language Models | Accepted to ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid development of AI-generated content (AIGC) technology has led to the misuse of highly realistic AI-generated images (AIGI) in spreading misinformation, posing a threat to public information security. Although existing AIGI detection techniques are generally effective, they face two issues: 1) a lack of human-verifiable explanations, and 2) a lack of generalization in the latest generation technology. To address these issues, we introduce a large-scale and comprehensive dataset, Holmes-Set, which includes the Holmes-SFTSet, an instruction-tuning dataset with explanations on whether images are AI-generated, and the Holmes-DPOSet, a human-aligned preference dataset. Our work introduces an efficient data annotation method called the Multi-Expert Jury, enhancing data generation through structured MLLM explanations and quality control via cross-model evaluation, expert defect filtering, and human preference modification. In addition, we propose Holmes Pipeline, a meticulously designed three-stage training framework comprising visual expert pre-training, supervised fine-tuning, and direct preference optimization. Holmes Pipeline adapts multimodal large language models (MLLMs) for AIGI detection while generating human-verifiable and human-aligned explanations, ultimately yielding our model AIGI-Holmes. During the inference stage, we introduce a collaborative decoding strategy that integrates the model perception of the visual expert with the semantic reasoning of MLLMs, further enhancing the generalization capabilities. Extensive experiments on three benchmarks validate the effectiveness of our AIGI-Holmes. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:26:31 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhou",
"Ziyin",
""
],
[
"Luo",
"Yunpeng",
""
],
[
"Wu",
"Yuanchen",
""
],
[
"Sun",
"Ke",
""
],
[
"Ji",
"Jiayi",
""
],
[
"Yan",
"Ke",
""
],
[
"Ding",
"Shouhong",
""
],
[
"Sun",
"Xiaoshuai",
""
],
[
"Wu",
"Yunsheng",
""
],
[
"Ji",
"Rongrong",
""
]
] |
2507.02665 | David Bernholdt | Timo Kehrer, Robert Haines, Guido Juckeland, Shurui Zhou, and David E. Bernholdt | Do Research Software Engineers and Software Engineering Researchers Speak the Same Language? | Early access journal version: T. Kehrer, R. Haines, G. Juckeland, S. Zhou and D. E. Bernholdt, "Do Research Software Engineers and Software Engineering Researchers Speak the Same Language?," in Computing in Science & Engineering, doi: 10.1109/MCSE.2025.3557236 | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Anecdotal evidence suggests that Research Software Engineers (RSEs) and Software Engineering Researchers (SERs) often use different terminologies for similar concepts, creating communication challenges. To better understand these divergences, we have started investigating how SE fundamentals from the SER community are interpreted within the RSE community, identifying aligned concepts, knowledge gaps, and areas for potential adaptation. Our preliminary findings reveal opportunities for mutual learning and collaboration, and our systematic methodology for terminology mapping provides a foundation for a crowd-sourced extension and validation in the future. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:27:54 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kehrer",
"Timo",
""
],
[
"Haines",
"Robert",
""
],
[
"Juckeland",
"Guido",
""
],
[
"Zhou",
"Shurui",
""
],
[
"Bernholdt",
"David E.",
""
]
] |
2507.02666 | Junyu Wang | Junyu Wang and Tianrui Wang and Meng Ge and Longbiao Wang and Jianwu Dang | ASDA: Audio Spectrogram Differential Attention Mechanism for Self-Supervised Representation Learning | Accepted at Interspeech2025 | null | null | null | cs.SD cs.AI cs.CL eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent advancements in audio self-supervised representation learning, the standard Transformer architecture has emerged as the predominant approach, yet its attention mechanism often allocates a portion of attention weights to irrelevant information, potentially impairing the model's discriminative ability. To address this, we introduce a differential attention mechanism, which effectively mitigates ineffective attention allocation through the integration of dual-softmax operations and appropriately tuned differential coefficients. Experimental results demonstrate that our ASDA model achieves state-of-the-art (SOTA) performance across multiple benchmarks, including audio classification (49.0% mAP on AS-2M, 41.5% mAP on AS20K), keyword spotting (98.3% accuracy on SPC-2), and environmental sound classification (96.1% accuracy on ESC-50). These results highlight ASDA's effectiveness in audio tasks, paving the way for broader applications. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:29:43 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Wang",
"Junyu",
""
],
[
"Wang",
"Tianrui",
""
],
[
"Ge",
"Meng",
""
],
[
"Wang",
"Longbiao",
""
],
[
"Dang",
"Jianwu",
""
]
] |
2507.02667 | Steven Giacalone | Steven Giacalone, Andrew W. Howard, Ryan A. Rubenzahl, Fei Dai, Luke B. Handley, Howard Isaacson, Samuel Halverson, Max Brodheim, Matt Brown, Theron W. Carmichael, William Deich, Benjamin J. Fulton, Steven R. Gibson, Grant M. Hill, Bradford Holden, Aaron Householder, Russ R. Laher, Kyle Lanclos, Joel Payne, Erik A. Petigura, Arpita Roy, Christian Schwab, Martin M. Sirk, Josh Walawender | A Hot Jupiter with a Retrograde Orbit around a Sun-like Star and a Toy Model of Hot Jupiters in Wide Binary Star Systems | null | null | null | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | We report an observation of a transit of the hot Jupiter (HJ) KELT-23A b with the Keck Planet Finder spectrograph and a measurement of the sky-projected obliquity ($\lambda$) of its Sun-like ($T_{\rm eff} \approx 5900$ K) host star. We measured a projected stellar obliquity of $\lambda \approx 180^\circ$, indicating that the orbit of the HJ is retrograde relative to the direction of the stellar spin. Due to the slow sky-projected rotational velocity of the host star ($v \sin{i_\star} \approx 0.5$ km s$^{-1}$), the true orbit of the HJ could be closer to polar. HJs around stars with effective temperatures below the Kraft break -- such as KELT-23A -- are generally found to have prograde orbits that are well-aligned with the equatorial planes of their host stars (i.e., $\lambda \sim 0^\circ$), most likely due to spin-orbit realignment driven by stellar tidal dissipation. This system is therefore a unique outlier that strains migration and tidal theories. The fact that the HJ has a highly misaligned orbit may suggest that the planet arrived at its close-in orbit relatively recently, possibly via interactions with the wide-separation (570 AU) M-dwarf companion in the system, or that it has stalled near an antialigned or polar orientation while realigning. Using Gaia DR3, we determined the orbit of the stellar companion to be moderately face-on ($\gamma = 60 \pm 4^\circ$). We show that the distribution of observed systems in the $\gamma - \lambda$ plane can be broadly reproduced using a toy model in which the orbits of the planetary and stellar companions begin aligned with the equatorial plane of the primary star and, upon migrating inwards, the planet preferentially obtains either an aligned or polar orbit. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:30:11 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Giacalone",
"Steven",
""
],
[
"Howard",
"Andrew W.",
""
],
[
"Rubenzahl",
"Ryan A.",
""
],
[
"Dai",
"Fei",
""
],
[
"Handley",
"Luke B.",
""
],
[
"Isaacson",
"Howard",
""
],
[
"Halverson",
"Samuel",
""
],
[
"Brodheim",
"Max",
""
],
[
"Brown",
"Matt",
""
],
[
"Carmichael",
"Theron W.",
""
],
[
"Deich",
"William",
""
],
[
"Fulton",
"Benjamin J.",
""
],
[
"Gibson",
"Steven R.",
""
],
[
"Hill",
"Grant M.",
""
],
[
"Holden",
"Bradford",
""
],
[
"Householder",
"Aaron",
""
],
[
"Laher",
"Russ R.",
""
],
[
"Lanclos",
"Kyle",
""
],
[
"Payne",
"Joel",
""
],
[
"Petigura",
"Erik A.",
""
],
[
"Roy",
"Arpita",
""
],
[
"Schwab",
"Christian",
""
],
[
"Sirk",
"Martin M.",
""
],
[
"Walawender",
"Josh",
""
]
] |
2507.02668 | Zhe Yee Tan | Zhe Yee Tan | MEGANet-W: A Wavelet-Driven Edge-Guided Attention Framework for Weak Boundary Polyp Detection | 7 pages, 3 figures | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Colorectal polyp segmentation is critical for early detection of colorectal cancer, yet weak and low contrast boundaries significantly limit automated accuracy. Existing deep models either blur fine edge details or rely on handcrafted filters that perform poorly under variable imaging conditions. We propose MEGANet-W, a Wavelet Driven Edge Guided Attention Network that injects directional, parameter free Haar wavelet edge maps into each decoder stage to recalibrate semantic features. Our two main contributions are: (1) a two-level Haar wavelet head for multi orientation edge extraction; and (2) Wavelet Edge Guided Attention (WEGA) modules that fuse wavelet cues with reverse and input branches. On five public polyp datasets, MEGANetW consistently outperforms existing methods, improving mIoU by up to 2.3% and mDice by 1.2%, while introducing no additional learnable parameters. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:31:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Tan",
"Zhe Yee",
""
]
] |
2507.02669 | Suman Pal | Suman Pal and Gargi Chaudhuri | Density dependent speed of sound and its consequences in neutron stars | null | Physics Letters B , Published, Available online 1 July 2025 | 10.1016/j.physletb.2025.139701 | null | nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a parametrized density-dependent speed of sound and construct an ensemble of equations of state for neutron stars which are found to closely resemble the realistic equations of state calculated using relativistic mean field theory. We show that each of these parameters display an unique feature relevant to the properties of the compact stars. The emergence of special points in the Mass-Radius plot is a significant outcome for neutron stars which is more commonly seen in case of hybrid stars. We have also shown that the curvature term in the speed of sound changes its sign for these hadronic equations of state without the matter reaching the conformal limit or undergoing any phase transition. It is related to the 1st derivative of the energy per nucleon reaching a maximum. We have also examined the detailed behavior of the trace anomaly and polytropic index for RMF models, as well as for a density-dependent parametrized speed of sound. Our analysis demonstrates that the sign of the trace anomaly at high densities is sensitive to the stiffness or softness of the EOS. Different observational constraints from mass-radius and tidal deformability can restrict the range of parameters in the proposed speed of sound model. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:33:58 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Pal",
"Suman",
""
],
[
"Chaudhuri",
"Gargi",
""
]
] |
2507.02670 | Lood Van Niekerk | Siqi Zhao, Joshua Moller, Porfi Quintero-Cadena, Lood van Niekerk | Guided Generation for Developable Antibodies | Published in ICML 2025 GenBio Workshop | null | null | null | cs.LG q-bio.BM | http://creativecommons.org/licenses/by/4.0/ | Therapeutic antibodies require not only high-affinity target engagement, but also favorable manufacturability, stability, and safety profiles for clinical effectiveness. These properties are collectively called `developability'. To enable a computational framework for optimizing antibody sequences for favorable developability, we introduce a guided discrete diffusion model trained on natural paired heavy- and light-chain sequences from the Observed Antibody Space (OAS) and quantitative developability measurements for 246 clinical-stage antibodies. To steer generation toward biophysically viable candidates, we integrate a Soft Value-based Decoding in Diffusion (SVDD) Module that biases sampling without compromising naturalness. In unconstrained sampling, our model reproduces global features of both the natural repertoire and approved therapeutics, and under SVDD guidance we achieve significant enrichment in predicted developability scores over unguided baselines. When combined with high-throughput developability assays, this framework enables an iterative, ML-driven pipeline for designing antibodies that satisfy binding and biophysical criteria in tandem. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:35:14 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhao",
"Siqi",
""
],
[
"Moller",
"Joshua",
""
],
[
"Quintero-Cadena",
"Porfi",
""
],
[
"van Niekerk",
"Lood",
""
]
] |
2507.02671 | Francesco Di Salvo | Francesco Di Salvo, Hanh Huyen My Nguyen, Christian Ledig | Embedding-Based Federated Data Sharing via Differentially Private Conditional VAEs | Accepted to MICCAI 2025 | null | null | null | cs.LG cs.CV eess.IV | http://creativecommons.org/licenses/by/4.0/ | Deep Learning (DL) has revolutionized medical imaging, yet its adoption is constrained by data scarcity and privacy regulations, limiting access to diverse datasets. Federated Learning (FL) enables decentralized training but suffers from high communication costs and is often restricted to a single downstream task, reducing flexibility. We propose a data-sharing method via Differentially Private (DP) generative models. By adopting foundation models, we extract compact, informative embeddings, reducing redundancy and lowering computational overhead. Clients collaboratively train a Differentially Private Conditional Variational Autoencoder (DP-CVAE) to model a global, privacy-aware data distribution, supporting diverse downstream tasks. Our approach, validated across multiple feature extractors, enhances privacy, scalability, and efficiency, outperforming traditional FL classifiers while ensuring differential privacy. Additionally, DP-CVAE produces higher-fidelity embeddings than DP-CGAN while requiring $5{\times}$ fewer parameters. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:36:15 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Di Salvo",
"Francesco",
""
],
[
"Nguyen",
"Hanh Huyen My",
""
],
[
"Ledig",
"Christian",
""
]
] |
2507.02672 | Qingyu Fan | Qingyu Fan, Yinghao Cai, Chao Li, Chunting Jiao, Xudong Zheng, Tao Lu, Bin Liang, Shuo Wang | MISCGrasp: Leveraging Multiple Integrated Scales and Contrastive Learning for Enhanced Volumetric Grasping | IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2025 | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robotic grasping faces challenges in adapting to objects with varying shapes and sizes. In this paper, we introduce MISCGrasp, a volumetric grasping method that integrates multi-scale feature extraction with contrastive feature enhancement for self-adaptive grasping. We propose a query-based interaction between high-level and low-level features through the Insight Transformer, while the Empower Transformer selectively attends to the highest-level features, which synergistically strikes a balance between focusing on fine geometric details and overall geometric structures. Furthermore, MISCGrasp utilizes multi-scale contrastive learning to exploit similarities among positive grasp samples, ensuring consistency across multi-scale features. Extensive experiments in both simulated and real-world environments demonstrate that MISCGrasp outperforms baseline and variant methods in tabletop decluttering tasks. More details are available at https://miscgrasp.github.io/. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:36:45 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fan",
"Qingyu",
""
],
[
"Cai",
"Yinghao",
""
],
[
"Li",
"Chao",
""
],
[
"Jiao",
"Chunting",
""
],
[
"Zheng",
"Xudong",
""
],
[
"Lu",
"Tao",
""
],
[
"Liang",
"Bin",
""
],
[
"Wang",
"Shuo",
""
]
] |
2507.02673 | Jonas Heimerl | Jonas Heimerl, Stefan Meier, Anne Herzig, Felix L\'opez Hoffmann, Lennart Seiffert, Daniel Lesko, Simon Hillmann, Simon Wittigschlager, Tobias Weitz, Thomas Fennel and Peter Hommelhoff | Attosecond physics in optical near fields | null | null | null | null | physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Attosecond science, the electron control by the field of ultrashort laser pulses, is maturing into lightfield-driven electronics, called petahertz electronics. Based on optical field-driven nanostructures, elements for petahertz electronics have been demonstrated. These hinge on the understanding of the electron dynamics in the optical near field of the nanostructure. Here we show near field-induced low energy stripes (NILES) in carrier-envelope phase-dependent electron spectra, a new spectral feature appearing in the direct electrons emitted from a strongly driven nanostructure, i.e., in the easily accessible energy region between 0 and a few electron volts. NILES emerge due to the sub-cycle sensitivity of ponderomotive acceleration of electrons injected into a strong near field gradient by a few-cycle optical waveform. NILES enables us to track the emission of direct and re-scattered electrons down to sub-cycle time-scales and to infer the electron momentum width at emission. Because NILES shows up in the direct part of the electrons, a large fraction of the emitted electrons can now be steered in new ways, facilitating the isolation of individual electron bursts with high charge density of 430 attosecond duration. These results not only substantially advance the understanding of attosecond physics in optical near fields, but also provide new ways of electron control for the nascent field of petahertz electronics. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:38:27 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Heimerl",
"Jonas",
""
],
[
"Meier",
"Stefan",
""
],
[
"Herzig",
"Anne",
""
],
[
"Hoffmann",
"Felix López",
""
],
[
"Seiffert",
"Lennart",
""
],
[
"Lesko",
"Daniel",
""
],
[
"Hillmann",
"Simon",
""
],
[
"Wittigschlager",
"Simon",
""
],
[
"Weitz",
"Tobias",
""
],
[
"Fennel",
"Thomas",
""
],
[
"Hommelhoff",
"Peter",
""
]
] |
2507.02674 | Tom Kneiphof | Tom Kneiphof and Reinhard Klein | Real-time Image-based Lighting of Glints | null | null | 10.1111/cgf.70175 | null | cs.GR cs.CV | http://creativecommons.org/licenses/by/4.0/ | Image-based lighting is a widely used technique to reproduce shading under real-world lighting conditions, especially in real-time rendering applications. A particularly challenging scenario involves materials exhibiting a sparkling or glittering appearance, caused by discrete microfacets scattered across their surface. In this paper, we propose an efficient approximation for image-based lighting of glints, enabling fully dynamic material properties and environment maps. Our novel approach is grounded in real-time glint rendering under area light illumination and employs standard environment map filtering techniques. Crucially, our environment map filtering process is sufficiently fast to be executed on a per-frame basis. Our method assumes that the environment map is partitioned into few homogeneous regions of constant radiance. By filtering the corresponding indicator functions with the normal distribution function, we obtain the probabilities for individual microfacets to reflect light from each region. During shading, these probabilities are utilized to hierarchically sample a multinomial distribution, facilitated by our novel dual-gated Gaussian approximation of binomial distributions. We validate that our real-time approximation is close to ground-truth renderings for a range of material properties and lighting conditions, and demonstrate robust and stable performance, with little overhead over rendering glints from a single directional light. Compared to rendering smooth materials without glints, our approach requires twice as much memory to store the prefiltered environment map. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:38:37 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kneiphof",
"Tom",
""
],
[
"Klein",
"Reinhard",
""
]
] |
2507.02675 | Zhaoqilin Yang | Zhaoqilin Yang, Xin Wang, Ruichen Zhang, Chanchan Li, Youliang Tian | TUC-PPO: Team Utility-Constrained Proximal Policy Optimization for Spatial Public Goods Games | null | null | null | null | cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce Team Utility-Constrained Proximal Policy Optimization (TUC-PPO), a new deep reinforcement learning framework. It extends Proximal Policy Optimization (PPO) by integrating team welfare objectives specifically for spatial public goods games. Unlike conventional approaches where cooperation emerges indirectly from individual rewards, TUC-PPO instead optimizes a bi-level objective integrating policy gradients and team utility constraints. Consequently, all policy updates explicitly incorporate collective payoff thresholds. The framework preserves PPO's policy gradient core while incorporating constrained optimization through adaptive Lagrangian multipliers. Therefore, decentralized agents dynamically balance selfish and cooperative incentives. The comparative analysis demonstrates superior performance of this constrained deep reinforcement learning approach compared to unmodified PPO and evolutionary game theory baselines. It achieves faster convergence to cooperative equilibria and greater stability against invasion by defectors. The framework formally integrates team objectives into policy updates. This work advances multi-agent deep reinforcement learning for social dilemmas while providing new computational tools for evolutionary game theory research. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:38:42 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yang",
"Zhaoqilin",
""
],
[
"Wang",
"Xin",
""
],
[
"Zhang",
"Ruichen",
""
],
[
"Li",
"Chanchan",
""
],
[
"Tian",
"Youliang",
""
]
] |
2507.02676 | Wendelin Werner | Amelia Carpenter and Wendelin Werner | On Loops in critical high-dimensional percolation | null | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We discuss the following type of results about critical Bernoulli percolation in high dimensions: The collection of clusters that do contain large (self-avoiding) loops in a large box is tight. The collection of these large loops has scaling limits that one will be able to relate to Brownian loop-soups (each of these atypical loop-containing clusters does in some sense only contain one large self-avoiding loop in the sense that any two such loops will be very close). This feature contrasts with the known proliferation of "typical" percolation clusters (i.e., among the many large clusters in a given box, only a handful will contain a loop of size comparable to the box). | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:38:47 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Carpenter",
"Amelia",
""
],
[
"Werner",
"Wendelin",
""
]
] |
2507.02677 | Kang Liu | Kang Liu and Enrique Zuazua | Moments, Time-Inversion and Source Identification for the Heat Equation | 30 pages | null | null | null | math.NA cs.NA math.AP math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We address the initial source identification problem for the heat equation, a notably ill-posed inverse problem characterized by exponential instability. Departing from classical Tikhonov regularization, we propose a novel approach based on moment analysis of the heat flow, transforming the problem into a more stable inverse moment formulation. By evolving the measured terminal time moments backward through their governing ODE system, we recover the moments of the initial distribution. We then reconstruct the source by solving a convex optimization problem that minimizes the total variation of a measure subject to these moment constraints. This formulation naturally promotes sparsity, yielding atomic solutions that are sums of Dirac measures. Compared to existing methods, our moment-based approach reduces exponential error growth to polynomial growth with respect to the terminal time. We provide explicit error estimates on the recovered initial distributions in terms of moment order, terminal time, and measurement errors. In addition, we develop efficient numerical discretization schemes and demonstrate significant stability improvements of our approach through comprehensive numerical experiments. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:39:31 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Kang",
""
],
[
"Zuazua",
"Enrique",
""
]
] |
2507.02678 | Camilla Ancona | Camilla Ancona, Dora Ricci, Carmela Bernardo, Francesco Lo Iudice, Anton Proskurnikov and Francesco Vasca | Imitation and Heterogeneity Shape the Resilience of Community Currency Networks | null | null | null | null | cs.CE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Community currency networks are made up of individuals and or companies that share some physical or social characteristics and engage in economic transactions using a virtual currency. This paper investigates the structural and dynamic properties of such mutual credit systems through a case study of Sardex, a community currency initiated and mainly operating in Sardinia, Italy. The transaction network is modeled as a directed weighted graph and analyzed through a graph theoretic framework focused on the analysis of strongly connected components, condensed representations, and behavioral connectivity patterns. Emphasis is placed on understanding the evolution of the network's core and peripheral structures over a three year period, with attention to temporal contraction, flow asymmetries, and structural fragmentation depending on different user types. Our findings reveal persistent deviations from degree based null models and suggest the presence of behavioral imitation, specifically, a user preference for more active peers. We further assess the impact of heterogeneous connections between different type of users, which strengthen the network topology and enhance its resilience. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:40:08 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ancona",
"Camilla",
""
],
[
"Ricci",
"Dora",
""
],
[
"Bernardo",
"Carmela",
""
],
[
"Iudice",
"Francesco Lo",
""
],
[
"Proskurnikov",
"Anton",
""
],
[
"Vasca",
"Francesco",
""
]
] |
2507.02679 | Ahmed Sabir | Ahmed Sabir and Rajesh Sharama | Exploring Gender Bias Beyond Occupational Titles | Work in progress | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | In this work, we investigate the correlation between gender and contextual biases, focusing on elements such as action verbs, object nouns, and particularly on occupations. We introduce a novel dataset, GenderLexicon, and a framework that can estimate contextual bias and its related gender bias. Our model can interpret the bias with a score and thus improve the explainability of gender bias. Also, our findings confirm the existence of gender biases beyond occupational stereotypes. To validate our approach and demonstrate its effectiveness, we conduct evaluations on five diverse datasets, including a Japanese dataset. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 14:42:03 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Sabir",
"Ahmed",
""
],
[
"Sharama",
"Rajesh",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.