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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2412.15784 | Mikael Santonen | Mikael Santonen, Antti Lahti, Zahra Jahanshah Rad, Mikko Miettinen, Masoud Ebrahimzadeh, Juha-Pekka Lehti\"o, Enni Snellman, Pekka Laukkanen, Marko Punkkinen, Kalevi Kokko, Katja Parkkinen, Markus Eklund | A detailed examination of polysilicon resistivity incorporating the grain size distribution | 7 pages, 7 figures, 2 tables | IEEE Transactions on Electron Devices, Volume: 72, Issue: 3, March 2025 | 10.1109/TED.2025.3530865 | null | cond-mat.mtrl-sci physics.comp-ph | http://creativecommons.org/licenses/by/4.0/ | Current transport in polysilicon is a complicated process with many factors to consider. The inhomogeneous nature of polysilicon with its differently shaped and sized grains is one such consideration. We have developed a method that enhances existing resistivity models with a two-dimensional extension that incorporates the grain size distribution using a Voronoi-based resistor network. We obtain grain size distributions both from our growth simulations (700 K, 800 K, and 900 K) and experimental analysis. Applying our method, we investigate the effect that variation in grain size produces with cases of different average grain sizes (2 nm to 3 $\mu$m). For example, the resistivity of polysilicon with an average grain size of 175 nm drops from 11 k$\Omega$ $\cdot$ cm to 4.5 k$\Omega$ $\cdot$ cm when compared to conventional one-dimensional modeling. Our study highlights the strong effect of grain size variation on resistivity, revealing that wider distributions result in significant resistivity reductions of up to more than 50%. Due to the larger grains present with a grain size distribution, current transport encounters fewer grain boundaries while the average grain size remains the same resulting in fewer barriers along the current transport path. Incorporating the grain structure into the resistivity modeling facilitates a more detailed and comprehensive characterization of the electrical properties of polysilicon. | [
{
"version": "v1",
"created": "Fri, 20 Dec 2024 10:52:53 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Santonen",
"Mikael",
""
],
[
"Lahti",
"Antti",
""
],
[
"Rad",
"Zahra Jahanshah",
""
],
[
"Miettinen",
"Mikko",
""
],
[
"Ebrahimzadeh",
"Masoud",
""
],
[
"Lehtiö",
"Juha-Pekka",
""
],
[
"Snellman",
"Enni",
""
],
[
"Laukkanen",
"Pekka",
""
],
[
"Punkkinen",
"Marko",
""
],
[
"Kokko",
"Kalevi",
""
],
[
"Parkkinen",
"Katja",
""
],
[
"Eklund",
"Markus",
""
]
] |
2412.15969 | Djalil Chafai | Djalil Chafa\"i and Max Fathi | On cutoff via rigidity for high dimensional curved diffusions | 17 pages, reworked, added Fisher information, added heat kernel, added product condition | null | null | null | math.PR math.DG math.FA | http://creativecommons.org/licenses/by/4.0/ | We consider overdamped Lanvegin diffusions in Euclidean space, with curvature equal to the spectral gap. This includes the Ornstein-Uhlenbeck process as well as non Gaussian and non product extensions with convex interaction, such as the Dyson process from random matrix theory. We show that a cutoff phenomenon or abrupt convergence to equilibrium occurs in high-dimension, at a critical time equal to the logarithm of the dimension divided by twice the spectral gap. This cutoff holds for Wasserstein distance, total variation, relative entropy, and Fisher information. A key observation is a relation to a spectral rigidity, linked to the presence of a Gaussian factor. A novelty is the extensive usage of functional inequalities, even for short-time regularization, and the reduction to Wasserstein. The proofs are short and conceptual. Since the product condition is satisfied, an Lp cutoff hold for all p. We moreover discuss a natural extension to Riemannian manifolds, a link with logarithmic gradient estimates in short-time for the heat kernel, and ask about stability by perturbation. Finally, beyond rigidity but still for diffusions, a discussion around the recent progresses on the product condition for nonnegatively curved diffusions leads us to introduce a new curvature product condition. | [
{
"version": "v1",
"created": "Fri, 20 Dec 2024 15:13:45 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:27:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chafaï",
"Djalil",
""
],
[
"Fathi",
"Max",
""
]
] |
2412.17556 | Marcelo E. Rubio | Juan Manuel Diaz, Marcelo E. Rubio | Bekenstein bounds in maximally symmetric nonlinear electrodynamics | 20 pages, 6 figures. Accepted for publication in PRD | null | null | null | gr-qc | http://creativecommons.org/licenses/by/4.0/ | We explore dynamical features of the maximally symmetric nonlinear extension of classical electromagnetism, recently proposed in the literature as ``ModMax'' electrodynamics. This family of theories is the only one that preserves all the symmetries of Maxwell's theory, having applications in the study of regular black hole solutions and supersymmetry. The purpose of this article is three-fold. Firstly, we study the initial-value problem of ModMax and show, by means of a simple geometric criterion, that such a theory admits a well-posed formulation. Secondly, we prove a series of geometric inequalities relating energy, charge, angular momentum, and size in ModMax. The validity of these bounds gives strong evidence of an universal inequality conjectured by Bekenstein for macroscopic systems. Finally, we perform the first stable numerical simulations of ModMax in the highly nonlinear regime. We apply our simulations for verifying an inequality between energy, size and angular momentum in bounded domains, and for showing birefringence as a function of the nonlinear parameter, comparing with other nonlinear nonlinear extensions. | [
{
"version": "v1",
"created": "Mon, 23 Dec 2024 13:22:00 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:17:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Diaz",
"Juan Manuel",
""
],
[
"Rubio",
"Marcelo E.",
""
]
] |
2412.17616 | Yanjie Dong | Zixuan Shangguan and Yanjie Dong and Song Guo and Victor C. M. Leung and M. Jamal Deen and Xiping Hu | Facial Expression Analysis and Its Potentials in IoT Systems: A Contemporary Survey | null | null | 10.1145/3737456 | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Facial expressions convey human emotions and can be categorized into macro-expressions (MaEs) and micro-expressions (MiEs) based on duration and intensity. While MaEs are voluntary and easily recognized, MiEs are involuntary, rapid, and can reveal concealed emotions. The integration of facial expression analysis with Internet-of-Thing (IoT) systems has significant potential across diverse scenarios. IoT-enhanced MaE analysis enables real-time monitoring of patient emotions, facilitating improved mental health care in smart healthcare. Similarly, IoT-based MiE detection enhances surveillance accuracy and threat detection in smart security. Our work aims to provide a comprehensive overview of research progress in facial expression analysis and explores its potential integration with IoT systems. We discuss the distinctions between our work and existing surveys, elaborate on advancements in MaE and MiE analysis techniques across various learning paradigms, and examine their potential applications in IoT. We highlight challenges and future directions for the convergence of facial expression-based technologies and IoT systems, aiming to foster innovation in this domain. By presenting recent developments and practical applications, our work offers a systematic understanding of the ways of facial expression analysis to enhance IoT systems in healthcare, security, and beyond. | [
{
"version": "v1",
"created": "Mon, 23 Dec 2024 14:41:01 GMT"
},
{
"version": "v2",
"created": "Thu, 22 May 2025 06:11:04 GMT"
},
{
"version": "v3",
"created": "Sat, 24 May 2025 03:00:40 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Shangguan",
"Zixuan",
""
],
[
"Dong",
"Yanjie",
""
],
[
"Guo",
"Song",
""
],
[
"Leung",
"Victor C. M.",
""
],
[
"Deen",
"M. Jamal",
""
],
[
"Hu",
"Xiping",
""
]
] |
2412.18486 | Mark Whitmeyer | Mark Whitmeyer | Calibrating the Subjective | null | null | null | null | econ.TH | http://creativecommons.org/licenses/by/4.0/ | I conduct Rabin's (2000) calibration exercise in the subjective expected utility realm. I show that the rejection of some risky bet by a risk-averse agent only implies the rejection of more extreme and less desirable bets and nothing more. | [
{
"version": "v1",
"created": "Tue, 24 Dec 2024 15:18:27 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:18:02 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Whitmeyer",
"Mark",
""
]
] |
2412.19032 | Nobuyoshi Komatsu | Nobuyoshi Komatsu | Cosmological scenario based on the first and second laws of thermodynamics: Thermodynamic constraints on a generalized cosmological model | Final version accepted for publication in EPJC. Several paragraphs and references are revised. [11 pages, 1 figure] | Eur. Phys. J. C 85, 607 (2025) | 10.1140/epjc/s10052-025-14332-5 | null | gr-qc astro-ph.CO hep-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The first and second laws of thermodynamics should lead to a consistent scenario for discussing the cosmological constant problem. In the present study, to establish such a thermodynamic scenario, cosmological equations in a flat Friedmann-Lema\^{i}tre-Robertson-Walker universe were derived from the first law, using an arbitrary entropy $S_{H}$ on a cosmological horizon. Then, the cosmological equations were formulated based on a general formulation that includes two extra driving terms, $f_{\Lambda}(t)$ and $h_{\textrm{B}}(t)$, which are usually used for, e.g., time-varying $\Lambda (t)$ cosmology and bulk viscous cosmology, respectively. In addition, thermodynamic constraints on the two terms are examined using the second law of thermodynamics, extending a previous analysis [Phys. Rev. D 99, 043523 (2019) (arXiv:1810.11138)]. It is found that a deviation $S_{\Delta}$ of $S_{H}$ from the Bekenstein-Hawking entropy plays important roles in the two terms. The second law should constrain the upper limits of $f_{\Lambda}(t)$ and $h_{\textrm{B}}(t)$ in our late Universe. The orders of the two terms are likely consistent with the order of the cosmological constant $\Lambda_{\textrm{obs}}$ measured by observations. In particular, when the deviation $S_{\Delta}$ is close to zero, $h_{\textrm{B}}(t)$ and $f_{\Lambda}(t)$ should reduce to zero and a constant value (consistent with the order of $\Lambda_{\textrm{obs}}$), respectively, as if a consistent and viable scenario could be obtained from thermodynamics. | [
{
"version": "v1",
"created": "Thu, 26 Dec 2024 03:04:19 GMT"
},
{
"version": "v2",
"created": "Wed, 05 Mar 2025 06:07:24 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 02:38:35 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Komatsu",
"Nobuyoshi",
""
]
] |
2412.19499 | Muhammad Nadeem | Syeda Amina Shabbir, Frank Fei Yun, Muhammad Nadeem, and Xiaolin Wang | Tailoring Robust Quantum Anomalous Hall Effect via Entropy-Engineering | null | Adv. Mater. 2025, 2503319 | 10.1002/adma.202503319 | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | The development of quantum materials and the tailoring of their functional properties is of fundamental interest in materials science. Here, a new design concept is proposed for the robust quantum anomalous Hall effect via entropy engineering in 2D magnets. As a prototypical example, the configurational entropy of monolayer transition metal trihalide VCl$_3$ is manipulated by incorporating four different transition-metal cations [Ti,Cr,Fe,Co] into the honeycomb structure made of vanadium, such that all in-plane mirror symmetries, inversion and/or roto-inversion are broken. Monolayer VCl$_3$ is a ferromagnetic Dirac half-metal in which spin-polarized Dirac dispersion at valley momenta is accompanied by bulk states at the $\Gamma$-point and thus the spin-orbit interaction-driven quantum anomalous Hall phase does not exhibit fully gapped bulk band dispersion. Entropy-driven bandstructure renormalization, especially band flattening in combination with red- and blue-shifts at different momenta of the Brillouin zone and crystal-field effects, transforms Dirac half-metal to a Dirac spin-gapless semiconductor and leads to a robust quantum anomalous Hall phase with fully gapped bulk band dispersion and, thus, a purely topological edge state transport without mixing with dissipative bulk channels. These findings provide a paradigm for designing entropy-engineered 2D materials for the realization of robust quantum anomalous Hall effect and quantum device applications. | [
{
"version": "v1",
"created": "Fri, 27 Dec 2024 07:33:57 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 23:54:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Shabbir",
"Syeda Amina",
""
],
[
"Yun",
"Frank Fei",
""
],
[
"Nadeem",
"Muhammad",
""
],
[
"Wang",
"Xiaolin",
""
]
] |
2412.20442 | Toru Kojo | Toru Kojo | Stiffening of matter in quark-hadron continuity: a mini-review | 7 pages, 5 figures; presentation given at "Aspects of Criticality II," Wroclaw, Poland; v2 revised, 8 pages, 5 figures | null | null | null | nucl-th astro-ph.HE hep-ph | http://creativecommons.org/licenses/by/4.0/ | Recent observations of neutron stars, combined with causality, thermodynamic stability, and nuclear constraints, indicate a rapid stiffening of QCD matter at densities slightly above nuclear saturation density ($n_0 \simeq 0.16\,{\rm fm}^{-3}$). The evolution of the stiffening is faster than expected from purely nucleonic models with many-body repulsion. Taking into account the quark substructure of baryons, we argue that the saturation of quark states occurs at $\sim$ 2-3$n_0$, driving quark matter formation even before baryonic cores of radius $\sim$0.5 fm spatially overlap. We describe the continuous transitions from hadronic to quark matter within a quarkyonic matter model in which gluons are assumed to remain confining at densities of interest. To obtain analytic insight into the transient regime, we construct an ideal model of quarkyonic matter, the {\it IdylliQ} model, in which one can freely switch from baryonic to quark descriptions and vice versa. | [
{
"version": "v1",
"created": "Sun, 29 Dec 2024 11:52:26 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:14:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kojo",
"Toru",
""
]
] |
2412.20938 | Hai-Zhou Lu | Yiyuan Chen, Xiaoxiong Liu, Hai-Zhou Lu, X. C. Xie | Electrical switching of altermagnetism | 8 pages, 5 figures | null | null | null | cond-mat.mes-hall cond-mat.mtrl-sci physics.app-ph | http://creativecommons.org/licenses/by/4.0/ | Switching magnetism using only electricity is of great significance for industrial applications but remains challenging. We find that, altermagnetism, as a newly discovered unconventional magnetism, may open an avenue along this effort. Specifically, to have deterministic switching, i.e., reversing current direction must reverse magnetic structure, parity symmetry has to be broken. We discover that, due to their symmetry which depends on chemical environments, altermagnet devices may naturally carry the parity symmetry breaking required for deterministic electrical switching of magnetism. More importantly, we identify MnTe bilayers (Te-Mn-Te-Mn-Te) as candidate devices, with the help of symmetry analysis, first-principles calculations, and magnetic dynamics simulations. This scheme will inspire further explorations on unconventional magnetism. | [
{
"version": "v1",
"created": "Mon, 30 Dec 2024 13:29:02 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 14:24:56 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chen",
"Yiyuan",
""
],
[
"Liu",
"Xiaoxiong",
""
],
[
"Lu",
"Hai-Zhou",
""
],
[
"Xie",
"X. C.",
""
]
] |
2412.21006 | Joonwon Jang | Joonwon Jang, Jaehee Kim, Wonbin Kweon, Seonghyeon Lee, Hwanjo Yu | Verbosity-Aware Rationale Reduction: Effective Reduction of Redundant Rationale via Principled Criteria | ACL 2025 FINDINGS | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) rely on generating extensive intermediate reasoning units (e.g., tokens, sentences) to enhance final answer quality across a wide range of complex tasks. While this approach has proven effective, it inevitably increases substantial inference costs. Previous methods adopting token-level reduction without clear criteria result in poor performance compared to models trained with complete rationale. To address this challenge, we propose a novel sentence-level rationale reduction framework leveraging likelihood-based criteria, verbosity, to identify and remove redundant reasoning sentences. Unlike previous approaches, our method leverages verbosity to selectively remove redundant reasoning sentences while preserving reasoning capabilities. Our experimental results across various reasoning tasks demonstrate that our method improves performance by an average of 7.71% while reducing token generation by 19.87% compared to model trained with complete reasoning paths. | [
{
"version": "v1",
"created": "Mon, 30 Dec 2024 15:15:08 GMT"
},
{
"version": "v2",
"created": "Tue, 31 Dec 2024 03:06:15 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 01:28:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jang",
"Joonwon",
""
],
[
"Kim",
"Jaehee",
""
],
[
"Kweon",
"Wonbin",
""
],
[
"Lee",
"Seonghyeon",
""
],
[
"Yu",
"Hwanjo",
""
]
] |
2501.01299 | Harshith Gowrachari | Harshith Gowrachari, Giovanni Stabile and Gianluigi Rozza | Model Reduction for Transport-Dominated Problems via Cross-Correlation Based Snapshot Registration | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional linear approximation methods, such as proper orthogonal decomposition and the reduced basis method, are ill-suited for transport-dominated problems due to the slow decay of the Kolmogorov $n$-width, leading to inefficient and inaccurate reduced-order models. In this work, we propose a model reduction approach for transport-dominated problems by employing cross-correlation based snapshot registration to accelerate the Kolmogorov $n$-width decay, thereby enabling the construction of efficient and accurate reduced-order models using linear approximation methods. We propose a complete framework comprising offline-online stages for the development of reduced order models using the cross-correlation based snapshots registration. The effectiveness of the proposed approach is demonstrated using two test cases: 1D travelling waves and the higher-order methods benchmark test case, 2D isentropic convective vortex. | [
{
"version": "v1",
"created": "Thu, 02 Jan 2025 15:26:49 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:13:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gowrachari",
"Harshith",
""
],
[
"Stabile",
"Giovanni",
""
],
[
"Rozza",
"Gianluigi",
""
]
] |
2501.01444 | Zhenhua Shi | Mingyue Guo, Zhenhua Shi | Local isometric immersions of pseudospherical surfaces described by a class of third order differential equations | This submission has been superseded by arXiv:2505.19728. Please refer to and cite the updated version | null | null | null | math.DG math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We discuss a specific type of pseudospherical surfaces defined by a class of third order differential equations, of the form $u_t - u_{xxt} = \lambda u^2 u_{xxx} + G(u, u_x, u_{xx})$, and poses a question about the dependence of the triples $\{a,b,c\}$ of the second fundamental form in the context of local isometric immersion in $\mathbb{E}^3$. It is demonstrated that the triples $\{a,b,c\}$ of the second fundamental form are not influenced by a jet of finite order of $u$. Instead, they are shown to rely on a jet of order zero, making them universal and not reliant on the specific solution chosen for $u$. | [
{
"version": "v1",
"created": "Sat, 28 Dec 2024 05:49:48 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:39:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Guo",
"Mingyue",
""
],
[
"Shi",
"Zhenhua",
""
]
] |
2501.01817 | Huiming Li | Huiming Li, Hao Chen, Xiangke Wang, Zhongkui Li and Lincheng Shen | Distributed Framework Construction for Affine Formation Control | second round of review | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In affine formation control problems, the construction of the framework with universal rigidity and affine localizability is a critical prerequisite, but it has not yet been well addressed, especially when additional agents join the formation or link/agent failures emerge. Motivated by this observation, we investigate the problem of constructing affine frameworks in three scenarios, including vertex addition, edge deletion and vertex deletion. Our approach starts from the original affine formation and uses geometric methods to locally adjust the structure of the weighted graph to describe the topology, so that the modified framework maintains the universal rigidity and affine localizability. Notably, the developed strategies only utilize local measurements and exhibit distributed characteristics, laying the foundation for applications in multi-agent systems. To demonstrate the compatibility with affine formation control proposals, we present a case study on affine formation tracking in a multi-UAV formation, demonstrating the effectiveness of our algorithms in constructing eligible frameworks in aforementioned scenarios. Moreover, a comparative simulation is also conducted to highlight the low time complexity of our distributed algorithm relative to the centralized optimization-based method. | [
{
"version": "v1",
"created": "Fri, 03 Jan 2025 14:02:37 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 14:01:43 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Huiming",
""
],
[
"Chen",
"Hao",
""
],
[
"Wang",
"Xiangke",
""
],
[
"Li",
"Zhongkui",
""
],
[
"Shen",
"Lincheng",
""
]
] |
2501.01965 | Andrea Tosin | Martina Fraia, Nadia Loy, Andrea Tosin | Linear inelastic kinetic equations modelling the spread of fake news and its interplay with personal awareness | 31 pages, 6 figures | Kinet. Relat. Models, 2025 | 10.3934/krm.2025013 | null | physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we introduce a kinetic model which describes a learning process leading individuals to build personal awareness about fake news. Next, we embed the results of this model into another kinetic model, which describes the popularity gained by news on social media conditioned to the reliability of the disseminated information. Both models are formulated in terms of linear inelastic Boltzmann-type equations, of which we investigate the main analytical properties - existence and uniqueness of solutions, trend to equilibrium, identification of the equilibrium distributions - by employing extensively Fourier methods for kinetic equations. We also provide evidence of the analytical results by means of Monte Carlo numerical simulations. | [
{
"version": "v1",
"created": "Wed, 18 Dec 2024 13:53:18 GMT"
},
{
"version": "v2",
"created": "Fri, 30 May 2025 08:56:45 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Fraia",
"Martina",
""
],
[
"Loy",
"Nadia",
""
],
[
"Tosin",
"Andrea",
""
]
] |
2501.02131 | William O'Regan | William O'Regan | Sum-product phenomena for Ahlfors-regular sets | 10 pages | null | null | null | math.CA math.CO math.MG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We utilise the recent work of Orponen to yield a result of sum and product for Ahlfors-regular sets. The result is sharp up to constants. As a corollary, we obtain the fractal analogue of Solymosi's $4/3$-bound for finite subsets of $\mathbb{R}.$ | [
{
"version": "v1",
"created": "Fri, 03 Jan 2025 22:51:02 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:36:52 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"O'Regan",
"William",
""
]
] |
2501.02423 | Shuaipeng Li | Xingwu Sun, Shuaipeng Li, Ruobing Xie, Weidong Han, Kan Wu, Zhen Yang, Yixing Li, An Wang, Shuai Li, Jinbao Xue, Yu Cheng, Yangyu Tao, Zhanhui Kang, Chengzhong Xu, Di Wang, Jie Jiang | Scaling Laws for Floating Point Quantization Training | null | null | null | null | cs.LG cs.AR cs.CL | http://creativecommons.org/licenses/by/4.0/ | Low-precision training is considered an effective strategy for reducing both training and downstream inference costs. Previous scaling laws for precision mainly focus on integer quantization, which pay less attention to the constituents in floating-point (FP) quantization, and thus cannot well fit the LLM losses in this scenario. In contrast, while FP quantization training is more commonly implemented in production, it's research has been relatively superficial. In this paper, we thoroughly explore the effects of FP quantization targets, exponent bits, mantissa bits, and the calculation granularity of the scaling factor in FP quantization training performance of LLM models. In addition to an accurate FP quantization unified scaling law, we also provide valuable suggestions for the community: (1) Exponent bits contribute slightly more to the model performance than mantissa bits. We provide the optimal exponent-mantissa bit ratio for different bit numbers, which is available for future reference by hardware manufacturers; (2) We discover the formation of the critical data size in low-precision LLM training. Too much training data exceeding the critical data size will inversely bring in degradation of LLM performance; (3) The optimal FP quantization precision is directly proportional to the computational power, but within a wide computational power range. We estimate that the best cost-performance precision should lie between 4-8 bits. | [
{
"version": "v1",
"created": "Sun, 05 Jan 2025 02:30:41 GMT"
},
{
"version": "v2",
"created": "Tue, 13 May 2025 13:19:32 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 09:06:06 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Sun",
"Xingwu",
""
],
[
"Li",
"Shuaipeng",
""
],
[
"Xie",
"Ruobing",
""
],
[
"Han",
"Weidong",
""
],
[
"Wu",
"Kan",
""
],
[
"Yang",
"Zhen",
""
],
[
"Li",
"Yixing",
""
],
[
"Wang",
"An",
""
],
[
"Li",
"Shuai",
""
],
[
"Xue",
"Jinbao",
""
],
[
"Cheng",
"Yu",
""
],
[
"Tao",
"Yangyu",
""
],
[
"Kang",
"Zhanhui",
""
],
[
"Xu",
"Chengzhong",
""
],
[
"Wang",
"Di",
""
],
[
"Jiang",
"Jie",
""
]
] |
2501.04250 | Ajay Singh | Ajay Singh and Trevor Brown | Publish on Ping: A Better Way to Publish Reservations in Memory Reclamation for Concurrent Data Structures | Extended version of full paper accepted at PPoPP '25: The 30th ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming Proceedings | null | null | null | cs.DC cs.PL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Safe memory reclamation techniques that utilize per read reservations, such as hazard pointers, often cause significant overhead in traversals of linked concurrent data structures. This is primarily due to the need to announce a reservation, and fence to enforce appropriate ordering, before each read. In read-intensive workloads, this overhead is amplified because, even if relatively little memory reclamation actually occurs, the full overhead of reserving records is still incurred while traversing data structures.
In this paper, we propose a novel memory reclamation technique by combining POSIX signals and delayed reclamation, introducing a publish-on-ping approach. This method eliminates the need to make reservations globally visible before use. Instead, threads privately track which records they are accessing, and share this information on demand with threads that intend to reclaim memory. The approach can serve as a drop-in replacement for hazard pointers and hazard eras. Furthermore, the capability to retain reservations during traversals in data structure operations and publish them on demand facilitates the construction of a variant of hazard pointers (EpochPOP). This variant uses epochs to approach the performance of epoch-based reclamation in the common case where threads are not frequently delayed (while retaining the robustness of hazard pointers).
Our publish-on-ping implementations based on hazard pointers (HP) and hazard eras, when applied to various data structures, exhibit significant performance improvements. The improvements across various workloads and data structures range from 1.2X to 4X over the original HP, up to 20% compared to a heavily optimized HP implementation similar to the one in the Folly open-source library, and up to 3X faster than hazard eras. EpochPOP delivers performance similar to epoch-based reclamation while providing stronger guarantees. | [
{
"version": "v1",
"created": "Wed, 08 Jan 2025 03:18:41 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 11:59:33 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Singh",
"Ajay",
""
],
[
"Brown",
"Trevor",
""
]
] |
2501.05397 | Sergei Khlebnikov | Sergei Khlebnikov | Entropy flow in a parametric amplifier | 10 pages, 2 figures; published in Phys. Rev. A | Phys. Rev. A 111, 052210 (2025) | 10.1103/PhysRevA.111.052210 | null | quant-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Computations of entropy in thermodynamics rely on discreteness of the spectra of the subsystems. We argue that, for cases with continuous spectra (typically, radiation), there is a useful definition of entropy flow based on discretizing the signal into Gabor's "atoms," say, by means of a windowed Fourier transform. In particular, applying this method to a parametric amplifier (paramp) driven by a classical pump and coupled to a zero-temperature Markovian bath, we find that the output entropy flux vanishes at large times, even though the energy and photon number fluxes remain nonzero. This is consistent with the manner in which the paramp is expected to release information about its initial state. We relate the quenching of the entropy flow to development of the off-diagonal coherences in the output and discuss possible relevance of this mechanism to the black-hole information problem. We also propose to use measurements of the off-diagonal coherences as a means of extracting the rates at which high-entropy subsystems release information to linear environments. | [
{
"version": "v1",
"created": "Thu, 09 Jan 2025 17:44:40 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Jan 2025 18:49:33 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 17:01:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Khlebnikov",
"Sergei",
""
]
] |
2501.05790 | Taywon Min | Taywon Min and Haeone Lee and Yongchan Kwon and Kimin Lee | Understanding Impact of Human Feedback via Influence Functions | Accepted at ACL 2025, Source code: https://github.com/mintaywon/IF_RLHF | null | null | null | cs.AI cs.HC cs.LG | http://creativecommons.org/licenses/by/4.0/ | In Reinforcement Learning from Human Feedback (RLHF), it is crucial to learn suitable reward models from human feedback to align large language models (LLMs) with human intentions. However, human feedback can often be noisy, inconsistent, or biased, especially when evaluating complex responses. Such feedback can lead to misaligned reward signals, potentially causing unintended side effects during the RLHF process. To address these challenges, we explore the use of influence functions to measure the impact of human feedback on the performance of reward models. We propose a compute-efficient approximation method that enables the application of influence functions to LLM-based reward models and large-scale preference datasets. In our experiments, we demonstrate two key applications of influence functions: (1) detecting common forms of labeler bias in human feedback datasets and (2) guiding labelers to refine their strategies to align more closely with expert feedback. By quantifying the impact of human feedback on reward models, we believe that influence functions can enhance feedback interpretability and contribute to scalable oversight in RLHF, helping labelers provide more accurate and consistent feedback. Source code is available at https://github.com/mintaywon/IF_RLHF | [
{
"version": "v1",
"created": "Fri, 10 Jan 2025 08:50:38 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:00:11 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Min",
"Taywon",
""
],
[
"Lee",
"Haeone",
""
],
[
"Kwon",
"Yongchan",
""
],
[
"Lee",
"Kimin",
""
]
] |
2501.05855 | Antonin Poche | Antonin Poch\'e (IRIT), Alon Jacovi, Agustin Martin Picard, Victor Boutin (CERCO UMR5549, ANITI), Fanny Jourdan | ConSim: Measuring Concept-Based Explanations' Effectiveness with Automated Simulatability | null | ACL 2025, Jul 2025, Vienna (Austria), France | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Concept-based explanations work by mapping complex model computations to human-understandable concepts. Evaluating such explanations is very difficult, as it includes not only the quality of the induced space of possible concepts but also how effectively the chosen concepts are communicated to users. Existing evaluation metrics often focus solely on the former, neglecting the latter. We introduce an evaluation framework for measuring concept explanations via automated simulatability: a simulator's ability to predict the explained model's outputs based on the provided explanations. This approach accounts for both the concept space and its interpretation in an end-to-end evaluation. Human studies for simulatability are notoriously difficult to enact, particularly at the scale of a wide, comprehensive empirical evaluation (which is the subject of this work). We propose using large language models (LLMs) as simulators to approximate the evaluation and report various analyses to make such approximations reliable. Our method allows for scalable and consistent evaluation across various models and datasets. We report a comprehensive empirical evaluation using this framework and show that LLMs provide consistent rankings of explanation methods. Code available at https://github.com/AnonymousConSim/ConSim. | [
{
"version": "v1",
"created": "Fri, 10 Jan 2025 10:53:48 GMT"
},
{
"version": "v2",
"created": "Mon, 13 Jan 2025 10:39:54 GMT"
},
{
"version": "v3",
"created": "Mon, 03 Feb 2025 09:13:17 GMT"
},
{
"version": "v4",
"created": "Wed, 04 Jun 2025 08:45:02 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Poché",
"Antonin",
"",
"IRIT"
],
[
"Jacovi",
"Alon",
"",
"CERCO UMR5549, ANITI"
],
[
"Picard",
"Agustin Martin",
"",
"CERCO UMR5549, ANITI"
],
[
"Boutin",
"Victor",
"",
"CERCO UMR5549, ANITI"
],
[
"Jourdan",
"Fanny",
""
]
] |
2501.06164 | Satchel Grant | Satchel Grant | Model Alignment Search | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | When can we say that two neural systems are the same? The answer to this question is goal-dependent, and it is often addressed through correlative methods such as Representational Similarity Analysis (RSA) and Centered Kernel Alignment (CKA). What nuances do we miss, however, when we fail to causally probe the representations? Do the dangers of cause vs. correlation exist in comparative representational analyses? In this work, we introduce a method for connecting neural representational similarity to behavior through causal interventions. The method learns orthogonal transformations that find an aligned subspace in which behavioral information from multiple distributed networks' representations can be isolated and interchanged. We first show that the method can be used to transfer the behavior from one frozen Neural Network (NN) to another in a manner similar to model stitching, and we show how the method can complement correlative similarity measures like RSA. We then introduce an efficient subspace orthogonalization technique using the Gram-Schmidt process -- that can also be used for Distributed Alignment Search (DAS) -- allowing us to perform analyses on larger models. Next, we empirically and theoretically show how our method can be equivalent to model stitching when desired, or it can take a form that is more restrictive to causal information, and in both cases, it reduces the number of required matrices for a comparison of n models from quadratic to linear in n. We then show how we can augment the loss objective with an auxiliary loss to train causally relevant alignments even when we can only read the representations from one of the two networks during training (like with biological networks). Lastly, we use number representations as a case study to explore how our method can be used to compare specific types of representational information across tasks and models. | [
{
"version": "v1",
"created": "Fri, 10 Jan 2025 18:39:29 GMT"
},
{
"version": "v2",
"created": "Thu, 16 Jan 2025 21:07:04 GMT"
},
{
"version": "v3",
"created": "Fri, 31 Jan 2025 10:07:41 GMT"
},
{
"version": "v4",
"created": "Thu, 24 Apr 2025 02:13:50 GMT"
},
{
"version": "v5",
"created": "Tue, 03 Jun 2025 19:11:11 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Grant",
"Satchel",
""
]
] |
2501.07093 | En-Jui Chang | En-Jui Chang | High-Rate Extended Binomial Codes for Multi-Qubit Encoding | 7 pages, 3 tables | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We introduce a new class of bosonic quantum error-correcting codes, termed \emph{extended binomial codes}, which generalize the structure of one-mode binomial codes by incorporating ideas from high-rate qubit stabilizer codes. These codes are constructed in close analogy to $[[n,k,d]]$ qubit codes, where the parameter $n$ corresponds to the total excitation budget rather than the number of physical qubits. Our construction achieves a significant reduction in average excitation per mode while preserving error-correcting capabilities, offering improved compatibility with hardware constraints in the strong-dispersive regime. We demonstrate that extended binomial codes not only reduce the mean excitation required for encoding but also simplify syndrome extraction and logical gate implementation, particularly the logical $\bar{X}$ operation. These advantages suggest that extended binomial codes offer a scalable and resource-efficient approach for bosonic quantum error correction. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2025 07:04:05 GMT"
},
{
"version": "v2",
"created": "Fri, 07 Mar 2025 08:56:20 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Mar 2025 06:49:40 GMT"
},
{
"version": "v4",
"created": "Wed, 04 Jun 2025 06:20:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chang",
"En-Jui",
""
]
] |
2501.07181 | Pascal Begout | Pascal B\'egout (IMT), Jes\'us Ildefonso D\'iaz | On the compactness of the support of solitary waves of the complex saturated nonlinear Schr{\"o}dinger equation and related problems | /j.physd.2024.134516\ | Physica D: Nonlinear Phenomena, 2025, 472, pp.134516 | 27E8;10.1016 | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the vectorial stationary Schr{\"o}dinger equation -$\Delta$u + a U + b u = F, with a saturated nonlinearity U = u/|u| and with some complex coefficients (a, b) $\in$ C 2 . Besides the existence and uniqueness of solutions for the Dirichlet and Neumann problems, we prove the compactness of the support of the solution, under suitable conditions on (a, b) and even when the source in the right hand side F (x) is not vanishing for large values of |x|. The proof of the compactness of the support uses a local energy method, given the impossibility of applying the maximum principle. We also consider the associate Schr{\"o}dinger-Poisson system when coupling with a simple magnetic field. Among other consequences, our results give a rigorous proof of the existence of ''solitons with compact support'' claimed, without any proof, by several previous authors. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2025 10:24:34 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:26:50 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bégout",
"Pascal",
"",
"IMT"
],
[
"Díaz",
"Jesús Ildefonso",
""
]
] |
2501.07198 | Jaehyeok Choi | Jaehyeok Choi, Eunwoo Lee | Large charge operators at large spin from relativistically rotating vortices | 34 pages + appendices, 2 figures | JHEP05(2025)087 | 10.1007/JHEP05(2025)087 | null | hep-th physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | We study the ground states of CFTs with a global $U(1)$ symmetry on $\mathbb{R}\times S^2$ in the regime of large charge $Q$ and large angular momentum $J$, using large charge EFT. We find that in the range $Q \ll J \ll Q^2$, the ground state solution is a superfluid densely populated with vortices rotating at a constant angular velocity $\Omega$. This is a relativistic generalization of the known (non-relativistic) rigid rotation phase, which corresponds to the small $\Omega$ limit of our solution. In the regime $Q^{3/2}\ll J\ll Q^2$, our solution achieves lower energy than previously identified states. In this regime, most of the vortices move near the speed of light, and we obtain the chiral fluctuation modes propagating at the speed of light. Interestingly, we find that our ground state can be interpreted as a zero temperature charged normal fluid rotating at a constant angular velocity $\Omega$. We rederive this solution purely from the fluid dynamics. Based on the (already established) applicability of fluid description to large non-supersymmetric extremal AdS black holes, we find that the boundary stress tensor and $U(1)$ current of extremal AdS Kerr-Newman black hole align with those of our solution. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2025 10:45:39 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 05:59:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Choi",
"Jaehyeok",
""
],
[
"Lee",
"Eunwoo",
""
]
] |
2501.07544 | Rapha\"el Clou\^atre | Rapha\"el Clou\^atre, Colin Krisko | $C^*$-supports and abnormalities of operator systems | 18 pages. Version 2: streamlined exposition, minor edits throughout | null | null | null | math.OA math.FA | http://creativecommons.org/licenses/by/4.0/ | Let $S$ be a concrete operator system represented on some Hilbert space $H$. A $C^*$-support of $S$ is the $C^*$-algebra generated (via the Choi--Effros product) by $S$ inside an injective operator system acting on $H$. By leveraging Hamana's theory, we show that such a $C^*$-support is unique precisely when $C^*(S)$ is contained in every copy of the injective envelope of $S$ that acts on $H$. Further, we demonstrate how the uniqueness of certain $C^*$-supports can be used to give new characterizations of the unique extension property for $*$-representations, as well as the hyperrigidity of $S$. In another direction, we utilize the collection of all $C^*$-supports of $S$ to describe the subspace generated by the so-called abnormalities of $S$, thereby complementing a result of Kakariadis. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2025 18:26:04 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 19:37:25 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Clouâtre",
"Raphaël",
""
],
[
"Krisko",
"Colin",
""
]
] |
2501.07930 | Thibaut Boissin | Thibaut Boissin (IRIT), Franck Mamalet, Thomas Fel, Agustin Martin Picard, Thomas Massena (IRIT), Mathieu Serrurier (IRIT) | An Adaptive Orthogonal Convolution Scheme for Efficient and Flexible CNN Architectures | null | International Conference on Machine Learning, 2025, Vancouver, Canada | null | null | cs.AI cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Orthogonal convolutional layers are valuable components in multiple areas of machine learning, such as adversarial robustness, normalizing flows, GANs, and Lipschitz-constrained models. Their ability to preserve norms and ensure stable gradient propagation makes them valuable for a large range of problems. Despite their promise, the deployment of orthogonal convolution in large-scale applications is a significant challenge due to computational overhead and limited support for modern features like strides, dilations, group convolutions, and transposed convolutions. In this paper, we introduce AOC (Adaptative Orthogonal Convolution), a scalable method that extends a previous method (BCOP), effectively overcoming existing limitations in the construction of orthogonal convolutions. This advancement unlocks the construction of architectures that were previously considered impractical. We demonstrate through our experiments that our method produces expressive models that become increasingly efficient as they scale. To foster further advancement, we provide an open-source python package implementing this method, called Orthogonium ( https://github.com/deel-ai/orthogonium ) . | [
{
"version": "v1",
"created": "Tue, 14 Jan 2025 08:32:12 GMT"
},
{
"version": "v2",
"created": "Thu, 20 Mar 2025 08:31:58 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 07:59:15 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Boissin",
"Thibaut",
"",
"IRIT"
],
[
"Mamalet",
"Franck",
"",
"IRIT"
],
[
"Fel",
"Thomas",
"",
"IRIT"
],
[
"Picard",
"Agustin Martin",
"",
"IRIT"
],
[
"Massena",
"Thomas",
"",
"IRIT"
],
[
"Serrurier",
"Mathieu",
"",
"IRIT"
]
] |
2501.09060 | Greta Toni | Greta Toni, Ghassem Gozaliasl, Matteo Maturi, Lauro Moscardini, Alexis Finoguenov, Gianluca Castignani, Fabrizio Gentile, Kaija Virolainen, Caitlin M. Casey, Jeyhan S. Kartaltepe, Hollis B. Akins, Natalie Allen, Rafael C. Arango-Toro, Arif Babul, Malte Brinch, Nicole E. Drakos, Andreas L. Faisst, Maximilien Franco, Richard E. Griffiths, Santosh Harish, G\"unther Hasinger, Olivier Ilbert, Shuowen Jin, Ali Ahmad Khostovan, Anton M. Koekemoer, Maarit Korpi-Lagg, Rebecca L. Larson, Jitrapon Lertprasertpong, Daizhong Liu, Georgios Magdis, Richard Massey, Henry J. McCracken, Jed McKinney, Louise Paquereau, Jason Rhodes, Brant E. Robertson, Mark Sargent, Marko Shuntov, Masayuki Tanaka, Sina Taamoli, Elmo Tempel, Sune Toft, Eleni Vardoulaki, Lilan Yang | The COSMOS-Web deep galaxy group catalog up to $z=3.7$ | 17 pages, 14 figures, 2 tables (including 2 appendices). Accepted for publication in A&A | A&A 697, A197 (2025) | 10.1051/0004-6361/202553759 | null | astro-ph.GA astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Galaxy groups with $M_{tot} \lesssim 10^{14}$ $M_\odot$ and up to a few tens of members are the most common galaxy environment, marking the transition between field and massive clusters. Identifying groups plays a crucial role in understanding structure formation and galaxy evolution. Modern deep surveys allow us to build well-characterized samples of groups up to the regime where structures were taking shape. We aimed to build the largest deep catalog of galaxy groups to date over the COSMOS-Web field effective area of 0.45 deg$^2$, leveraging the deep high quality data of the new COSMOS-Web photometric catalog resulted from the James Webb Space Telescope observations of the COSMOS-Web field. We performed the group search with the AMICO algorithm, a linear matched filter based on an analytical model for the group signal. AMICO has already been tested in wide and deep field surveys, including COSMOS data up to $z=2$. In this work, we tested the algorithm performances at even higher redshift and searched for protocluster cores at $z>2$. We compiled a list of known protoclusters in COSMOS at $2 \leq z \leq 3.7$, matched them with our detections and studied the clustering of the detected cores. We estimated purity and completeness of our sample by creating data-driven mocks with the SinFoniA code and linked signal-to-noise to purity. We detected 1678 groups in the COSMOS-Web field up to $z=3.7$, including lists of members extending nearly two magnitudes deeper than the previous AMICO-COSMOS catalog. 756 groups were detected with purity of 80\%. More than 500 groups have their redshift confirmed by assigning spectroscopic counterparts. This group catalog offers a unique opportunity to explore galaxy evolution in different environments spanning $\sim$12 Gyr and to study groups, from the least rich population to the formation of the most massive clusters. | [
{
"version": "v1",
"created": "Wed, 15 Jan 2025 19:00:00 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:22:30 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Toni",
"Greta",
""
],
[
"Gozaliasl",
"Ghassem",
""
],
[
"Maturi",
"Matteo",
""
],
[
"Moscardini",
"Lauro",
""
],
[
"Finoguenov",
"Alexis",
""
],
[
"Castignani",
"Gianluca",
""
],
[
"Gentile",
"Fabrizio",
""
],
[
"Virolainen",
"Kaija",
""
],
[
"Casey",
"Caitlin M.",
""
],
[
"Kartaltepe",
"Jeyhan S.",
""
],
[
"Akins",
"Hollis B.",
""
],
[
"Allen",
"Natalie",
""
],
[
"Arango-Toro",
"Rafael C.",
""
],
[
"Babul",
"Arif",
""
],
[
"Brinch",
"Malte",
""
],
[
"Drakos",
"Nicole E.",
""
],
[
"Faisst",
"Andreas L.",
""
],
[
"Franco",
"Maximilien",
""
],
[
"Griffiths",
"Richard E.",
""
],
[
"Harish",
"Santosh",
""
],
[
"Hasinger",
"Günther",
""
],
[
"Ilbert",
"Olivier",
""
],
[
"Jin",
"Shuowen",
""
],
[
"Khostovan",
"Ali Ahmad",
""
],
[
"Koekemoer",
"Anton M.",
""
],
[
"Korpi-Lagg",
"Maarit",
""
],
[
"Larson",
"Rebecca L.",
""
],
[
"Lertprasertpong",
"Jitrapon",
""
],
[
"Liu",
"Daizhong",
""
],
[
"Magdis",
"Georgios",
""
],
[
"Massey",
"Richard",
""
],
[
"McCracken",
"Henry J.",
""
],
[
"McKinney",
"Jed",
""
],
[
"Paquereau",
"Louise",
""
],
[
"Rhodes",
"Jason",
""
],
[
"Robertson",
"Brant E.",
""
],
[
"Sargent",
"Mark",
""
],
[
"Shuntov",
"Marko",
""
],
[
"Tanaka",
"Masayuki",
""
],
[
"Taamoli",
"Sina",
""
],
[
"Tempel",
"Elmo",
""
],
[
"Toft",
"Sune",
""
],
[
"Vardoulaki",
"Eleni",
""
],
[
"Yang",
"Lilan",
""
]
] |
2501.09255 | Zhuoyu Chen | Peng Li, Guangdi Zhou, Wei Lv, Yueying Li, Changming Yue, Haoliang Huang, Lizhi Xu, Jianchang Shen, Yu Miao, Wenhua Song, Zihao Nie, Yaqi Chen, Heng Wang, Weiqiang Chen, Yaobo Huang, Zhen-Hua Chen, Tian Qian, Junhao Lin, Junfeng He, Yu-Jie Sun, Zhuoyu Chen, Qi-Kun Xue | Angle-resolved photoemission spectroscopy of superconducting (La,Pr)3Ni2O7/SrLaAlO4 heterostructures | null | National Science Review, nwaf205 (2025) | 10.1093/nsr/nwaf205 | null | cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Ruddlesden-Popper bilayer nickelate thin film superconductors discovered under ambient pressure enable vast possibilities for investigating electronic structures of the superconducting state. Here, we report angle-resolved photoemission spectroscopy (ARPES) measurements of 1, 2, and 3 unit-cell epitaxial La2.85Pr0.15Ni2O7 films grown on SrLaAlO4 substates, through pure-oxygen in situ sample transportation. Evidence obtained using photons with distinct probing depths shows that conduction is localized primarily at the first unit cell near the interface. Scanning transmission electron microscopy (STEM), together with energy-dispersive X-ray spectroscopy (EDS) and electron energy loss spectroscopy (EELS), indicates that interfacial Sr diffusion and pronounced p-d hybridization gradient may collectively account for the interfacial confinement of conduction. Fermi surface maps reveal hole doping compared to non-superconducting ambient-pressure bulk crystals. Measurements of dispersive band structures suggest the contributions from both Ni dx2-y2 and dz2 orbitals at the Fermi level. Density functional theory (DFT) + U calculations capture qualitative features of the ARPES results, consistent with a hole-doped scenario. These findings constrain theoretical models of the superconducting mechanism and suggest potential for enhancing superconductivity in nickelates under ambient pressure. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 02:49:17 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:17:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Peng",
""
],
[
"Zhou",
"Guangdi",
""
],
[
"Lv",
"Wei",
""
],
[
"Li",
"Yueying",
""
],
[
"Yue",
"Changming",
""
],
[
"Huang",
"Haoliang",
""
],
[
"Xu",
"Lizhi",
""
],
[
"Shen",
"Jianchang",
""
],
[
"Miao",
"Yu",
""
],
[
"Song",
"Wenhua",
""
],
[
"Nie",
"Zihao",
""
],
[
"Chen",
"Yaqi",
""
],
[
"Wang",
"Heng",
""
],
[
"Chen",
"Weiqiang",
""
],
[
"Huang",
"Yaobo",
""
],
[
"Chen",
"Zhen-Hua",
""
],
[
"Qian",
"Tian",
""
],
[
"Lin",
"Junhao",
""
],
[
"He",
"Junfeng",
""
],
[
"Sun",
"Yu-Jie",
""
],
[
"Chen",
"Zhuoyu",
""
],
[
"Xue",
"Qi-Kun",
""
]
] |
2501.09328 | Yixiao Xu | Yixiao Xu, Binxing Fang, Rui Wang, Yinghai Zhou, Yuan Liu, Mohan Li, and Zhihong Tian | Neural Honeytrace: A Robust Plug-and-Play Watermarking Framework against Model Extraction Attacks | null | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Developing high-performance deep learning models is resource-intensive, leading model owners to utilize Machine Learning as a Service (MLaaS) platforms instead of publicly releasing their models. However, malicious users may exploit query interfaces to execute model extraction attacks, reconstructing the target model's functionality locally. While prior research has investigated triggerable watermarking techniques for asserting ownership, existing methods face significant challenges: (1) most approaches require additional training, resulting in high overhead and limited flexibility, and (2) they often fail to account for advanced attackers, leaving them vulnerable to adaptive attacks.
In this paper, we propose Neural Honeytrace, a robust plug-and-play watermarking framework against model extraction attacks. We first formulate a watermark transmission model from an information-theoretic perspective, providing an interpretable account of the principles and limitations of existing triggerable watermarking. Guided by the model, we further introduce: (1) a similarity-based training-free watermarking method for plug-and-play and flexible watermarking, and (2) a distribution-based multi-step watermark information transmission strategy for robust watermarking. Comprehensive experiments on four datasets demonstrate that Neural Honeytrace outperforms previous methods in efficiency and resisting adaptive attacks. Neural Honeytrace reduces the average number of samples required for a worst-case t-Test-based copyright claim from 193,252 to 1,857 with zero training cost. The code is available at https://github.com/NeurHT/NeurHT. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 06:59:20 GMT"
},
{
"version": "v2",
"created": "Fri, 17 Jan 2025 06:50:23 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 02:14:47 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Xu",
"Yixiao",
""
],
[
"Fang",
"Binxing",
""
],
[
"Wang",
"Rui",
""
],
[
"Zhou",
"Yinghai",
""
],
[
"Liu",
"Yuan",
""
],
[
"Li",
"Mohan",
""
],
[
"Tian",
"Zhihong",
""
]
] |
2501.09567 | Tong Liu | Tong Liu, Ying-Hai Wu, Hong-Hao Tu, Tao Xiang | Bridging conformal field theory and parton approaches to SU(n)_k chiral spin liquids | 15 pages, 6 figures | Phys. Rev. B 111, 205137 (2025) | 10.1103/PhysRevB.111.205137 | null | cond-mat.str-el cond-mat.stat-mech hep-th quant-ph | http://creativecommons.org/licenses/by/4.0/ | We employ the $\mathrm{SU}(n)_k$ Wess-Zumino-Witten (WZW) model in conformal field theory to construct lattice wave functions in both one and two dimensions. The spins on all lattice sites are chosen to transform under the $\mathrm{SU}(n)$ irreducible representation with a single row and $k$ boxes in the Young tableau. It is demonstrated that the wave functions can be reinterpreted as parton states, which enables efficient conversion to matrix product states such that many physical properties can be evaluated directly. In one dimension, these wave functions describe critical spin chains whose universality classes are in one-to-one correspondence with the WZW models used in the construction. In two dimensions, our constructions yield model wave functions for chiral spin liquids, and we show how to find all topological sectors of them in a systematic way. Using the null vectors of Kac-Moody algebras, parent Hamiltonians of the $\mathrm{SU}(3)_k$ series are derived. The $\mathrm{SU}(3)_k$ chiral spin liquids are lattice analogs of non-Abelian spin-singlet fractional quantum Hall states, and the $k=2$ member hosts Fibonacci anyons. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 14:42:00 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 06:22:21 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Liu",
"Tong",
""
],
[
"Wu",
"Ying-Hai",
""
],
[
"Tu",
"Hong-Hao",
""
],
[
"Xiang",
"Tao",
""
]
] |
2501.09621 | Tehila Dahan | Tehila Dahan, Kfir Y. Levy | Weight for Robustness: A Comprehensive Approach towards Optimal Fault-Tolerant Asynchronous ML | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | We address the challenges of Byzantine-robust training in asynchronous distributed machine learning systems, aiming to enhance efficiency amid massive parallelization and heterogeneous computing resources. Asynchronous systems, marked by independently operating workers and intermittent updates, uniquely struggle with maintaining integrity against Byzantine failures, which encompass malicious or erroneous actions that disrupt learning. The inherent delays in such settings not only introduce additional bias to the system but also obscure the disruptions caused by Byzantine faults. To tackle these issues, we adapt the Byzantine framework to asynchronous dynamics by introducing a novel weighted robust aggregation framework. This allows for the extension of robust aggregators and a recent meta-aggregator to their weighted versions, mitigating the effects of delayed updates. By further incorporating a recent variance-reduction technique, we achieve an optimal convergence rate for the first time in an asynchronous Byzantine environment. Our methodology is rigorously validated through empirical and theoretical analysis, demonstrating its effectiveness in enhancing fault tolerance and optimizing performance in asynchronous ML systems. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 16:00:52 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 13:31:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dahan",
"Tehila",
""
],
[
"Levy",
"Kfir Y.",
""
]
] |
2501.09805 | Avalon Campbell-Cousins | Avalon Campbell-Cousins, Federica Guazzo, Mark Bastin, Mario A. Parra, Javier Escudero | Multiplex Nodal Modularity: A novel network metric for the regional analysis of amnestic mild cognitive impairment during a working memory binding task | 35 pages, 9 figures, this is to be submitted to PLOS ONE for publication | null | null | null | q-bio.NC cs.SI physics.bio-ph | http://creativecommons.org/licenses/by/4.0/ | Modularity is a well-established concept for assessing community structures in various single and multi-layer networks, including those in biological and social domains. Brain networks are known to exhibit community structure at local, meso, and global scale. However, modularity is limited as a metric to a global scale describing the overall strength of community structure, overlooking important variations in community structure at node level. To address this limitation, we extended modularity to individual nodes. This novel measure of nodal modularity (nQ) captures both mesoscale and local-scale changes in modularity. We hypothesized that nQ would illuminate granular changes in the brain due to diseases such as Alzheimer's disease (AD), which are known to disrupt the brain's modular structure. We explored nQ in multiplex networks of a visual short-term memory binding task in fMRI and DTI data in the early stages of AD. While limited by sample size, changes in nQ for individual regions of interest (ROIs) in our fMRI networks were predominantly observed in visual, limbic, and paralimbic systems in the brain, aligning with known AD trajectories and linked to amyloid-$\beta$ and tau deposition. Furthermore, observed changes in white-matter microstructure in our DTI networks in parietal and frontal regions may compliment studies of white-matter integrity in poor memory binders. Additionally, nQ clearly differentiated MCI from MCI converters indicating that nQ may be sensitive to this key turning point of AD. Our findings demonstrate the utility of nQ as a measure of localized group structure, providing novel insights into task and disease-related variability at the node level. Given the widespread application of modularity as a global measure, nQ represents a significant advancement, providing a granular measure of network organization applicable to a wide range of disciplines. | [
{
"version": "v1",
"created": "Thu, 16 Jan 2025 19:27:20 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:05:44 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Campbell-Cousins",
"Avalon",
""
],
[
"Guazzo",
"Federica",
""
],
[
"Bastin",
"Mark",
""
],
[
"Parra",
"Mario A.",
""
],
[
"Escudero",
"Javier",
""
]
] |
2501.09976 | Changze Lv | Changze Lv, Jingwen Xu, Yiyang Lu, Xiaohua Wang, Zhenghua Wang, Zhibo Xu, Di Yu, Xin Du, Xiaoqing Zheng, Xuanjing Huang | Dendritic Localized Learning: Toward Biologically Plausible Algorithm | null | null | null | null | cs.NE | http://creativecommons.org/licenses/by/4.0/ | Backpropagation is the foundational algorithm for training neural networks and a key driver of deep learning's success. However, its biological plausibility has been challenged due to three primary limitations: weight symmetry, reliance on global error signals, and the dual-phase nature of training, as highlighted by the existing literature. Although various alternative learning approaches have been proposed to address these issues, most either fail to satisfy all three criteria simultaneously or yield suboptimal results. Inspired by the dynamics and plasticity of pyramidal neurons, we propose Dendritic Localized Learning (DLL), a novel learning algorithm designed to overcome these challenges. Extensive empirical experiments demonstrate that DLL satisfies all three criteria of biological plausibility while achieving state-of-the-art performance among algorithms that meet these requirements. Furthermore, DLL exhibits strong generalization across a range of architectures, including MLPs, CNNs, and RNNs. These results, benchmarked against existing biologically plausible learning algorithms, offer valuable empirical insights for future research. We hope this study can inspire the development of new biologically plausible algorithms for training multilayer networks and advancing progress in both neuroscience and machine learning. Our code is available at https://github.com/Lvchangze/Dendritic-Localized-Learning. | [
{
"version": "v1",
"created": "Fri, 17 Jan 2025 06:35:20 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:35:08 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Lv",
"Changze",
""
],
[
"Xu",
"Jingwen",
""
],
[
"Lu",
"Yiyang",
""
],
[
"Wang",
"Xiaohua",
""
],
[
"Wang",
"Zhenghua",
""
],
[
"Xu",
"Zhibo",
""
],
[
"Yu",
"Di",
""
],
[
"Du",
"Xin",
""
],
[
"Zheng",
"Xiaoqing",
""
],
[
"Huang",
"Xuanjing",
""
]
] |
2501.11085 | C. W. J. Beenakker | C.W.J. Beenakker | Entropy and singular-value moments of products of truncated random unitary matrices | 6 pages, 4 figures | Phys.Rev.E 111, 064108 (2025) | 10.1103/PhysRevE.111.064108 | null | quant-ph cond-mat.stat-mech math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Products of truncated unitary matrices, independently and uniformly drawn from the unitary group, can be used to study universal aspects of monitored quantum circuits. The von Neumann entropy of the corresponding density matrix decreases with increasing length $L$ of the product chain, in a way that depends on the matrix dimension $N$ and the truncation depth $\delta N$. Here we study that dependence in the double-scaling limit $L,N\rightarrow\infty$, at fixed ratio $\tau=L\delta N/N$. The entropy reduction crosses over from a linear to a logarithmic dependence on $\tau$ when this parameter crosses unity. The central technical result is an expression for the singular-value moments of the matrix product in terms of the Erlang function from queueing theory. | [
{
"version": "v1",
"created": "Sun, 19 Jan 2025 15:46:08 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Beenakker",
"C. W. J.",
""
]
] |
2501.11454 | Akash Kundu Mr | Akash Kundu | Improving thermal state preparation of Sachdev-Ye-Kitaev model with reinforcement learning on quantum hardware | Accepted in Machine Learning: Science and Technology. Code at https://github.com/Aqasch/solving_SYK_model_with_RL | null | null | null | quant-ph cs.AI cs.LG hep-lat hep-th | http://creativecommons.org/licenses/by/4.0/ | The Sachdev-Ye-Kitaev (SYK) model, known for its strong quantum correlations and chaotic behavior, serves as a key platform for quantum gravity studies. However, variationally preparing thermal states on near-term quantum processors for large systems ($N>12$, where $N$ is the number of Majorana fermions) presents a significant challenge due to the rapid growth in the complexity of parameterized quantum circuits. This paper addresses this challenge by integrating reinforcement learning (RL) with convolutional neural networks, employing an iterative approach to optimize the quantum circuit and its parameters. The refinement process is guided by a composite reward signal derived from entropy and the expectation values of the SYK Hamiltonian. This approach reduces the number of CNOT gates by two orders of magnitude for systems $N\geq12$ compared to traditional methods like first-order Trotterization. We demonstrate the effectiveness of the RL framework in both noiseless and noisy quantum hardware environments, maintaining high accuracy in thermal state preparation. This work advances a scalable, RL-based framework with applications for quantum gravity studies and out-of-time-ordered thermal correlators computation in quantum many-body systems on near-term quantum hardware. The code is available at https://github.com/Aqasch/solving_SYK_model_with_RL. | [
{
"version": "v1",
"created": "Mon, 20 Jan 2025 12:41:17 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 10:57:15 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kundu",
"Akash",
""
]
] |
2501.11772 | Biagio De Simone | M. G. Dainotti, B. De Simone, A. Garg, K. Kohri, A. Bashyal, A. Aich, A. Mondal, S. Nagataki, G. Montani, T. Jareen, V. M. Jabir, S. Khanjani, M. Bogdan, N. Fraija, A. C. C. do E. S. Pedreira, R. H. Dejrah, A. Singh, M. Parakh, R. Mandal, K. Jarial, G. Lambiase, H. Sarkar | A New Master Supernovae Ia sample and the investigation of the $H_0$ tension | 45 pages, 8 figures, 7 tables. Accepted for publication on the Journal of High Energy Astrophysics [JHEAP_100405]. Current version matches the proofs | null | null | KEK-TH-2711, KEK-Cosmo-0378 | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern cosmological research still thoroughly debates the discrepancy between local probes and the Cosmic Microwave Background observations in the Hubble constant (\texorpdfstring{$H_0$}{H0}) measurements, ranging from 4 to 6$\sigma$. In the current study, we examine this tension using the Supernovae Ia (SNe Ia) data from the Pantheon, Pantheon+ (P+), Joint Lightcurve Analysis (JLA), and Dark Energy Survey, (DES) catalogs combined together into the so-called Master Sample. The sample contains 3714 SNe Ia, and is divided all of them into redshift-ordered bins. Three binning techniques are presented: the equi-population, the moving window (MW), and the equi-spacing in the \texorpdfstring{$\log-z$}{log-z}. We perform a Markov-Chain Monte Carlo analysis (MCMC) for each bin to determine the $H_0$ value, estimating it within the standard flat \texorpdfstring{$\Lambda$CDM}{LCDM} and the \texorpdfstring{$w_{0}w_{a}$CDM}{w0waCDM} models. These \texorpdfstring{$H_0$}{H0} values are then fitted with the following phenomenological function: \texorpdfstring{$\mathcal{H}_0(z) = \tilde{H}_0 / (1 + z)^\alpha$}{H0(z) = H0tilde / (1 + z)^alpha}, where \texorpdfstring{$\tilde{H}_0$}{H0tilde} is a free parameter representing \texorpdfstring{$\mathcal{H}_0(z)$}{H0(z)} fitted in \texorpdfstring{$z=0$}{z=0}, and \texorpdfstring{$\alpha$}{alpha} is the evolutionary parameter. Our results indicate a decreasing trend characterized by \texorpdfstring{$\alpha \sim 0.01$}{alpha ~ 0.01}, whose consistency with zero ranges from $1 \sigma$ in 5 cases to 1 case at 3 $\sigma$ and 11 cases at $> 3 \sigma$ in several samples and configurations. Such a trend in the SNe Ia catalogs could be due to evolution with redshift for the astrophysical variables or unveiled selection biases. Alternatively, intrinsic physics, possibly the \texorpdfstring{$f(R)$}{f(R)} theory of gravity, could be responsible for this trend. | [
{
"version": "v1",
"created": "Mon, 20 Jan 2025 22:31:05 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Jan 2025 15:32:31 GMT"
},
{
"version": "v3",
"created": "Fri, 24 Jan 2025 22:00:31 GMT"
},
{
"version": "v4",
"created": "Wed, 14 May 2025 17:43:22 GMT"
},
{
"version": "v5",
"created": "Wed, 04 Jun 2025 15:38:07 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dainotti",
"M. G.",
""
],
[
"De Simone",
"B.",
""
],
[
"Garg",
"A.",
""
],
[
"Kohri",
"K.",
""
],
[
"Bashyal",
"A.",
""
],
[
"Aich",
"A.",
""
],
[
"Mondal",
"A.",
""
],
[
"Nagataki",
"S.",
""
],
[
"Montani",
"G.",
""
],
[
"Jareen",
"T.",
""
],
[
"Jabir",
"V. M.",
""
],
[
"Khanjani",
"S.",
""
],
[
"Bogdan",
"M.",
""
],
[
"Fraija",
"N.",
""
],
[
"Pedreira",
"A. C. C. do E. S.",
""
],
[
"Dejrah",
"R. H.",
""
],
[
"Singh",
"A.",
""
],
[
"Parakh",
"M.",
""
],
[
"Mandal",
"R.",
""
],
[
"Jarial",
"K.",
""
],
[
"Lambiase",
"G.",
""
],
[
"Sarkar",
"H.",
""
]
] |
2501.12435 | Bernhard Putzer | Bernhard Putzer and Mathias S. Scheurer | Eliashberg Theory and Superfluid Stiffness of Band-Off-Diagonal Pairing in Twisted Graphene | null | null | 10.1103/PhysRevB.111.144513 | null | cond-mat.supr-con cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Recently, band-off-diagonal superconductivity has been proposed [Nat. Commun. 14, 7134 (2023)] as a candidate pairing state for twisted graphene systems. Based on mean-field theory, it was shown that it not only naturally emerges from both intervalley electron-phonon coupling and fluctuations of the nearby correlated insulator, but also exhibits nodal and gapped regimes as indicated by scanning tunneling microscopy experiments. Here we study band-off-diagonal pairing within Eliashberg theory. We show that despite the additional frequency dependence, the leading-order description of both intervalley coherent fluctuations or intervalley phonons exhibits a symmetry prohibiting admixture of an intraband component to the interband pairing state. It is found that even- and odd-frequency pairing mix, which originates from the reduced number of flavor degrees of freedom in the normal state. From analytic continuation, we obtain the electronic spectral function showing that, also within Eliashberg theory, the interband nature leads to an enhanced spectral weight below the order-parameter energy compared to band-diagonal pairing. Finally, we also study the superfluid stiffness of band-off-diagonal pairing, taking into account multi-band and quantum geometry effects. It is shown that for $s$-wave and chiral momentum dependencies, conventionally leading to fully gapped phases, an interband structure reduces the temperature scale below which the stiffness saturates. Depending on parameters, for the chiral state, this scale can even be suppressed all the way to zero temperature leading to a complex competition of multiple dispersive and geometrical contributions. Our results show that interband pairing might also be able to explain more recent stiffness measurements in the superconducting state of twisted multilayer graphene. | [
{
"version": "v1",
"created": "Tue, 21 Jan 2025 19:00:00 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Putzer",
"Bernhard",
""
],
[
"Scheurer",
"Mathias S.",
""
]
] |
2501.13149 | Jenny Wagner | Jenny Wagner and David Benisty | Galaxy infall models for arbitrary velocity directions | 15 pages, 11 figures; in press at Astronomy & Astrophysics | null | 10.1051/0004-6361/202553892 | null | astro-ph.GA astro-ph.CO astro-ph.IM gr-qc | http://creativecommons.org/licenses/by-nc-nd/4.0/ | For most galaxies in the cosmos, our knowledge of their motion is limited to line-of-sight velocities from redshift observations. To determine the radial velocity between two galaxies the minor and major infall models were established by Karachentsev & Kashibadze (2006). Regardless of the background cosmology, our derivations reveal that these infall models approximate the total radial velocity between two galaxies by two different projections employing different information about the system. For galaxies having small angular separations $\theta$, all infall models agree that the radial velocity is the difference of their line-of-sight components. Applying these models to ca. $500$ halos of the Illustris-3 simulation, we find the perpendicular and tangential velocity parts to be non-negligible for more than 90% of all, more than 5000 infalling subhalos. Thus, even for $\theta < 10$ deg, the infall-model velocities deviate from the true radial velocity. Only for 30% we found the true one lay between the minor and major infall velocity. However, the infall models yield robust upper and lower bounds to the true radial velocity dispersion. Observed under $\theta < 10$ deg the velocity dispersion inferred from the sole difference of line-of-sight velocity components even coincides with the true one, justifying this approach for high-redshift groups and clusters. Based on these findings, we predict the radial velocity dispersion of the M81-group from the minor infall model (upper bound) $\sigma_{\mathrm{r,min}} = (180 \pm 42)~\mbox{km}/\mbox{s}$, from the major infall model (lower bound) $\sigma_{\mathrm{r,maj}} = (142 \pm 64) ~\mbox{km}/\mbox{s}$ and $\sigma_\mathrm{r,\Delta v} = (99 \pm 36)~\mbox{km}/\mbox{s}$ from the line-of-sight-velocity difference. | [
{
"version": "v1",
"created": "Wed, 22 Jan 2025 19:00:00 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Apr 2025 15:28:20 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 09:37:43 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wagner",
"Jenny",
""
],
[
"Benisty",
"David",
""
]
] |
2501.13486 | Emir Syahreza Fadhilla | Emir Syahreza Fadhilla, M Shoufie Ukhtary, Ardian Nata Atmaja, Bobby Eka Gunara | Radial Stabilization of Magnetic Skyrmions Under Strong External Magnetic Field | 12 pages, 5 figures, single column format | null | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | The skyrmion number density, $q\equiv\vec{n}\cdot\left(\partial_x\vec{n}\times\partial_y\vec{n}\right)/(4\pi)$, is one of the key quantities that characterizes the topological properties of a magnetic skyrmion. In this work, we propose a model for a two-dimensional magnetic system with Hamiltonian that contains an interaction term proportional to $q^2$ which preserves inversion symmetry. The proposed $q^2$ term is also known as the Skyrme term and is a two-dimensional version of the well-known quartic term in models of three-dimensional Hopfions. In contrast with the usual exchange interaction, the $q^2$ term persists at the strong external magnetic field limit. Using the Landau-Lifshitz-Gilbert equation for micromagnetic calculations, we show that the minimum energy configuration of this model exhibits skyrmion properties. Furthermore, this configuration remains stable under small linear radially symmetric perturbations, and we demonstrate that the total energy of the system is bounded from below, ensuring that it remains above the vacuum energy. This implies a topologically protected configuration. Our model provides a framework for describing skyrmions in materials without broken inversion symmetry, particularly in systems subjected to strong external magnetic fields, where conventional exchange interactions are significantly weaker than the Zeeman effect. | [
{
"version": "v1",
"created": "Thu, 23 Jan 2025 09:06:31 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 03:49:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Fadhilla",
"Emir Syahreza",
""
],
[
"Ukhtary",
"M Shoufie",
""
],
[
"Atmaja",
"Ardian Nata",
""
],
[
"Gunara",
"Bobby Eka",
""
]
] |
2501.14419 | Elise Paznokas | Elise Paznokas | Non-Invertible $SO(2)$ Symmetry of 4d Maxwell from Continuous Gaugings | 21 pages + appendices, 2 figures; references added, minor revisions to appendix A, (published version) | J. High Energ. Phys. 2025, 14 (2025) | 10.1007/JHEP06(2025)014 | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | We describe the self-duality symmetries for 4d Maxwell theory at any value of the coupling $\tau$ via topological manipulations that include gauging continuous symmetries with flat connections. Moreover, we demonstrate that the $SL(2,\mathbb{Z})$ duality of Maxwell can be realized by trivial gauging operations. Using a non-compact symmetry topological field theory (symTFT) to encode continuous global symmetries of the boundary theory, we reproduce the symTFT for Maxwell and find within this framework condensation defects that implement the non-invertible $SO(2)$ self-duality symmetry. These defects are systematically constructed by higher gauging subsets of the bulk $\mathbb{R}\times \mathbb{R}$ symmetry with appropriate discrete torsion. | [
{
"version": "v1",
"created": "Fri, 24 Jan 2025 11:41:16 GMT"
},
{
"version": "v2",
"created": "Fri, 02 May 2025 10:01:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Paznokas",
"Elise",
""
]
] |
2501.14478 | Arend Moerman | A. Moerman, K. Karatsu, J. J. A. Baselmans, S. O. Dabironezare, S. Fujita, R. Huiting, K. Kohno, Y. Nishimura, F. Steenvoorde, T. Takekoshi, Y. Tamura, A. Taniguchi, S. J. C. Yates, B. R. Brandl, A. Endo | Alignment and optical verification of DESHIMA 2.0 at ASTE | 22 pages, 11 figures. Version accepted by JATIS | null | 10.1117/1.JATIS.11.2.025007 | null | astro-ph.IM | http://creativecommons.org/licenses/by/4.0/ | We developed, characterized, and verified an alignment procedure for the DESHIMA 2.0 instrument, an ultra wide-band spectrometer operating between 200--400 GHz, at the ASTE telescope. To this end, we mounted the warm optics, consisting of a modified Dragonian dual reflector system, on a motor controlled hexapod. Crucial in the alignment procedure is our sky chopper, which allows fast beam switching. It has a small entrance and exit aperture coupling to (cold) sky, which creates a measurable signal with respect to the warm cabin environment. By scanning the instrument beam across the entrance aperture of the sky chopper using the hexapod, we found the hexapod configuration that produced the lowest signal on our detectors, implying the beam is coupled fully to cold sky and not the warm cabin. We first characterized the alignment procedure in the laboratory, where we used a vat containing liquid nitrogen as the cold source behind the sky chopper. Then, we applied the alignment procedure to DESHIMA 2.0 at ASTE. We found that the alignment procedure significantly improved the aperture efficiency compared to previously reported values of the aperture efficiency of DESHIMA at ASTE, which indicates the veracity of the alignment procedure. | [
{
"version": "v1",
"created": "Fri, 24 Jan 2025 13:26:58 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:39:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Moerman",
"A.",
""
],
[
"Karatsu",
"K.",
""
],
[
"Baselmans",
"J. J. A.",
""
],
[
"Dabironezare",
"S. O.",
""
],
[
"Fujita",
"S.",
""
],
[
"Huiting",
"R.",
""
],
[
"Kohno",
"K.",
""
],
[
"Nishimura",
"Y.",
""
],
[
"Steenvoorde",
"F.",
""
],
[
"Takekoshi",
"T.",
""
],
[
"Tamura",
"Y.",
""
],
[
"Taniguchi",
"A.",
""
],
[
"Yates",
"S. J. C.",
""
],
[
"Brandl",
"B. R.",
""
],
[
"Endo",
"A.",
""
]
] |
2501.14755 | Daoyuan Chen | Daoyuan Chen, Yilun Huang, Xuchen Pan, Nana Jiang, Haibin Wang, Yilei Zhang, Ce Ge, Yushuo Chen, Wenhao Zhang, Zhijian Ma, Jun Huang, Wei Lin, Yaliang Li, Bolin Ding, Jingren Zhou | Data-Juicer 2.0: Cloud-Scale Adaptive Data Processing for and with Foundation Models | 34 pages, 10 figures, 3 tables | null | null | null | cs.DC cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The burgeoning field of foundation models necessitates advanced data processing mechanisms capable of harnessing vast and valuable data with various types used by these models. Nevertheless, the current landscape presents unique challenges that traditional data processing frameworks struggle to handle effectively, particularly in handling the complexity of multimodal data. In response, we present Data-Juicer 2.0, a data processing system backed by 100+ data processing operators spanning text, image, video, and audio modalities, supporting more critical tasks including data analysis, synthesis, annotation, and foundation model post-training. With seamless compatibility and dedicated optimization for popular dataset hubs like Hugging Face and computing engines like Ray, it improves upon its predecessor in terms of usability, efficiency, and programmability. It features an easily accessible user interface layer that supports decoupled Python interactions, RESTful APIs, and conversational commands. It contains a new runtime layer optimized for adaptive execution and management across varying dataset scales, processing demands, and computational environments, while hiding unnecessary system details. Extensive empirical evaluations demonstrate Data-Juicer 2.0's remarkable performance and scalability, highlighting its capability to efficiently process TB-level data with 10k+ CPU cores. The system is publicly available and has been widely adopted in diverse research fields and real-world products such as Alibaba Cloud PAI. We actively maintain it and share insights from practical feedback, with the goal of facilitating research and application of next-generation foundation models. | [
{
"version": "v1",
"created": "Mon, 23 Dec 2024 08:29:57 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 13:46:21 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chen",
"Daoyuan",
""
],
[
"Huang",
"Yilun",
""
],
[
"Pan",
"Xuchen",
""
],
[
"Jiang",
"Nana",
""
],
[
"Wang",
"Haibin",
""
],
[
"Zhang",
"Yilei",
""
],
[
"Ge",
"Ce",
""
],
[
"Chen",
"Yushuo",
""
],
[
"Zhang",
"Wenhao",
""
],
[
"Ma",
"Zhijian",
""
],
[
"Huang",
"Jun",
""
],
[
"Lin",
"Wei",
""
],
[
"Li",
"Yaliang",
""
],
[
"Ding",
"Bolin",
""
],
[
"Zhou",
"Jingren",
""
]
] |
2501.14965 | Stefan Strohauer | Stefan Strohauer, Fabian Wietschorke, Markus D\"oblinger, Christian Schmid, Stefanie Grotowski, Lucio Zugliani, Bj\"orn Jonas, Kai M\"uller, Jonathan J. Finley | Origin of performance enhancement of superconducting nanowire single-photon detectors by He-ion irradiation | 18 pages, 11 figures | APL Quantum 2, 026131 (2025) | 10.1063/5.0259935 | null | quant-ph cond-mat.mes-hall cond-mat.supr-con physics.ins-det physics.optics | http://creativecommons.org/licenses/by/4.0/ | Superconducting nanowire single-photon detectors (SNSPDs) are indispensable in fields such as quantum science and technology, astronomy, and biomedical imaging, where high detection efficiency, low dark count rates and high timing accuracy are required. Recently, helium (He) ion irradiation was shown to be a promising method to enhance SNSPD performance. Here, we study how changes in the underlying superconducting NbTiN film and the SiO2/Si substrate affect device performance. While irradiated and unirradiated NbTiN films show similar crystallinity, we observe He bubble formation below the SiO2/Si interface and an amorphization of the Si substrate. Both reduce the thermal conductance between the superconducting thin film and the substrate from 210 W/m^2/K^4 to 70 W/m^2/K^4 after irradiation with 2000 ions/nm^2. This effect, combined with the lateral straggle of He ions in the substrate, allows the modification of the superconductor-to-substrate thermal conductance of an SNSPD by selectively irradiating the regions around the nanowire. With this approach, we achieved an increased plateau width of saturating intrinsic detection efficiency of 9.8 uA compared to 3.7 uA after full irradiation. Moreover, the critical current remained similar to that of the unirradiated reference device (59 uA versus 60.1 uA), while full irradiation reduced it to 22.4 uA. Our results suggest that the irradiation-induced reduction of the thermal conductance significantly enhances SNSPD sensitivity, offering a novel approach to locally engineer substrate properties for improved detector performance. | [
{
"version": "v1",
"created": "Fri, 24 Jan 2025 22:55:14 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Strohauer",
"Stefan",
""
],
[
"Wietschorke",
"Fabian",
""
],
[
"Döblinger",
"Markus",
""
],
[
"Schmid",
"Christian",
""
],
[
"Grotowski",
"Stefanie",
""
],
[
"Zugliani",
"Lucio",
""
],
[
"Jonas",
"Björn",
""
],
[
"Müller",
"Kai",
""
],
[
"Finley",
"Jonathan J.",
""
]
] |
2501.15358 | Duong Anh-Tai Tran | Tran Duong Anh-Tai and Miguel A. Garc\'ia-March and Thomas Busch and Thom\'as Fogarty | Quantum correlations and spatial localization in trapped one-dimensional ultra-cold Bose-Bose-Bose mixtures | 22 pages, 12 figures | null | null | null | cond-mat.quant-gas quant-ph | http://creativecommons.org/licenses/by/4.0/ | We systematically investigate and illustrate the complete ground-state phase diagram for a one-dimensional, three-species mixture of a few repulsively interacting bosons trapped harmonically. To numerically obtain the solutions to the many-body Schr\"{o}dinger equation, we employ the improved Exact Diagonalization method [T. D. Anh-Tai {\it et al.}, SciPost Physics 15, 048 (2023)], which is capable of treating strongly-correlated few-body systems from first principles in an efficiently truncated Hilbert space. We present our comprehensive results for all possible combinations of intra- and interspecies interactions in the extreme limits that are either the ideal limit ($g=0$) or close to the hard-core limit ($g\to\infty$). These results show the emergence of unique ground-state properties related to correlations, coherence and spatial localization stemming from strongly repulsive interactions. | [
{
"version": "v1",
"created": "Sun, 26 Jan 2025 01:28:41 GMT"
},
{
"version": "v2",
"created": "Tue, 22 Apr 2025 16:02:03 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 14:10:05 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Anh-Tai",
"Tran Duong",
""
],
[
"García-March",
"Miguel A.",
""
],
[
"Busch",
"Thomas",
""
],
[
"Fogarty",
"Thomás",
""
]
] |
2501.15911 | Florian Hantke | Florian Hantke, Peter Snyder, Hamed Haddadi, Ben Stock | Web Execution Bundles: Reproducible, Accurate, and Archivable Web Measurements | In Proceedings of the 34rd USENIX Security Symposium, USENIX Security 2025 | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Recently, reproducibility has become a cornerstone in the security and privacy research community, including artifact evaluations and even a new symposium topic. However, Web measurements lack tools that can be reused across many measurement tasks without modification, while being robust to circumvention, and accurate across the wide range of behaviors in the Web. As a result, most measurement studies use custom tools and varied archival formats, each of unknown correctness and significant limitations, systematically affecting the research's accuracy and reproducibility.
To address these limitations, we present WebREC, a Web measurement tool that is, compared against the current state-of-the-art, accurate (i.e., correctly measures and attributes events not possible with existing tools), general (i.e., reusable without modification for a broad range of measurement tasks), and comprehensive (i.e., handling events from all relevant browser behaviors). We also present .web, an archival format for the accurate and reproducible measurement of a wide range of website behaviors. We empirically evaluate WebREC's accuracy by replicating well-known Web measurement studies and showing that WebREC's results more accurately match our baseline. We then assess if WebREC and .web succeed as general-purpose tools, which could be used to accomplish many Web measurement tasks without modification. We find that this is so: 70% of papers discussed in a 2024 web crawling SoK paper could be conducted using WebREC as is, and a larger number (48%) could be leveraged against .web archives without requiring any new crawling. | [
{
"version": "v1",
"created": "Mon, 27 Jan 2025 10:03:03 GMT"
},
{
"version": "v2",
"created": "Thu, 30 Jan 2025 13:20:01 GMT"
},
{
"version": "v3",
"created": "Wed, 05 Feb 2025 07:46:09 GMT"
},
{
"version": "v4",
"created": "Wed, 04 Jun 2025 10:03:51 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hantke",
"Florian",
""
],
[
"Snyder",
"Peter",
""
],
[
"Haddadi",
"Hamed",
""
],
[
"Stock",
"Ben",
""
]
] |
2501.16150 | Pascal Sager | Pascal J. Sager and Benjamin Meyer and Peng Yan and Rebekka von Wartburg-Kottler and Layan Etaiwi and Aref Enayati and Gabriel Nobel and Ahmed Abdulkadir and Benjamin F. Grewe and Thilo Stadelmann | A Comprehensive Survey of Agents for Computer Use: Foundations, Challenges, and Future Directions | null | null | null | null | cs.AI cs.HC cs.SY eess.SY | http://creativecommons.org/licenses/by/4.0/ | Agents for computer use (ACUs) are an emerging class of systems capable of executing complex tasks on digital devices - such as desktops, mobile phones, and web platforms - given instructions in natural language. These agents can automate tasks by controlling software via low-level actions like mouse clicks and touchscreen gestures. However, despite rapid progress, ACUs are not yet mature for everyday use.
In this survey, we investigate the state-of-the-art, trends, and research gaps in the development of practical ACUs. We provide a comprehensive review of the ACU landscape, introducing a unifying taxonomy spanning three dimensions: (I) the domain perspective, characterizing agent operating contexts; (II) the interaction perspective, describing observation modalities (e.g., screenshots, HTML) and action modalities (e.g., mouse, keyboard, code execution); and (III) the agent perspective, detailing how agents perceive, reason, and learn.
We review 87 ACUs and 33 datasets across foundation model-based and classical approaches through this taxonomy. Our analysis identifies six major research gaps: insufficient generalization, inefficient learning, limited planning, low task complexity in benchmarks, non-standardized evaluation, and a disconnect between research and practical conditions.
To address these gaps, we advocate for: (a) vision-based observations and low-level control to enhance generalization; (b) adaptive learning beyond static prompting; (c) effective planning and reasoning methods and models; (d) benchmarks that reflect real-world task complexity; (e) standardized evaluation based on task success; (f) aligning agent design with real-world deployment constraints.
Together, our taxonomy and analysis establish a foundation for advancing ACU research toward general-purpose agents for robust and scalable computer use. | [
{
"version": "v1",
"created": "Mon, 27 Jan 2025 15:44:02 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 10:30:14 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Sager",
"Pascal J.",
""
],
[
"Meyer",
"Benjamin",
""
],
[
"Yan",
"Peng",
""
],
[
"von Wartburg-Kottler",
"Rebekka",
""
],
[
"Etaiwi",
"Layan",
""
],
[
"Enayati",
"Aref",
""
],
[
"Nobel",
"Gabriel",
""
],
[
"Abdulkadir",
"Ahmed",
""
],
[
"Grewe",
"Benjamin F.",
""
],
[
"Stadelmann",
"Thilo",
""
]
] |
2501.16748 | Garima Chhikara | Garima Chhikara, Abhishek Kumar and Abhijnan Chakraborty | Through the Prism of Culture: Evaluating LLMs' Understanding of Indian Subcultures and Traditions | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large Language Models (LLMs) have shown remarkable advancements but also raise concerns about cultural bias, often reflecting dominant narratives at the expense of under-represented subcultures. In this study, we evaluate the capacity of LLMs to recognize and accurately respond to the Little Traditions within Indian society, encompassing localized cultural practices and subcultures such as caste, kinship, marriage, and religion. Through a series of case studies, we assess whether LLMs can balance the interplay between dominant Great Traditions and localized Little Traditions. We explore various prompting strategies and further investigate whether using prompts in regional languages enhances the models cultural sensitivity and response quality. Our findings reveal that while LLMs demonstrate an ability to articulate cultural nuances, they often struggle to apply this understanding in practical, context-specific scenarios. To the best of our knowledge, this is the first study to analyze LLMs engagement with Indian subcultures, offering critical insights into the challenges of embedding cultural diversity in AI systems. | [
{
"version": "v1",
"created": "Tue, 28 Jan 2025 06:58:25 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 17:05:12 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chhikara",
"Garima",
""
],
[
"Kumar",
"Abhishek",
""
],
[
"Chakraborty",
"Abhijnan",
""
]
] |
2501.16989 | Arnaud Amblard | Arnaud Amblard and Aur\'elien Drezet | From Hamilton-Jacobi to Bohm: Why the Wave Function Isn't Just Another Action | null | null | null | null | quant-ph physics.hist-ph | http://creativecommons.org/licenses/by/4.0/ | This paper examines the physical meaning of the wave function in Bohmian mechanics (BM), addressing the debate between causal and nomological interpretations. While BM postulates particles with definite trajectories guided by the wave function, the ontological status of the wave function itself remains contested. Critics of the causal interpretation argue that the wave function's high-dimensionality and lack of back-reaction disqualify it as a physical entity. Proponents of the nomological interpretation, drawing parallels to the classical Hamiltonian, propose that the wave function is a "law-like" entity. However, this view faces challenges, including reliance on speculative quantum gravity frameworks (e.g., the Wheeler-DeWitt equation) and conceptual ambiguities about the nature of "nomological entities". By systematically comparing BM to Hamilton-Jacobi theory, this paper highlights disanalogies between the wave function and the classical action function. These differences, particularly the wave function's dynamical necessity and irreducibility, support a sui generis interpretation, where the wave function represents a novel ontological category unique to quantum theory. The paper concludes that the wave function's role in BM resists classical analogies, demanding a metaphysical framework that accommodates its non-local, high-dimensional, and dynamically irreducible nature. | [
{
"version": "v1",
"created": "Tue, 28 Jan 2025 14:42:00 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 17:36:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Amblard",
"Arnaud",
""
],
[
"Drezet",
"Aurélien",
""
]
] |
2501.17533 | Alina Joch | Alina Joch and G\"otz S. Uhrig and Benedikt Fauseweh | Entanglement-informed Construction of Variational Quantum Circuits | 18 pages, 18 figures | Quantum Sci. Technol. 10, 035032 (2025) | 10.1088/2058-9565/addb6f | null | quant-ph cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Variational Quantum Eigensolver (VQE) is a promising tool for simulating ground states of quantum many-body systems on noisy quantum computers. Its effectiveness relies heavily on the ansatz, which must be both hardware-efficient for implementation on noisy hardware and problem-specific to avoid local minima and convergence problems. In this article, we explore entanglement-informed ansatz schemes that naturally emerge from specific models, aiming to balance accuracy with minimal use of two-qubit entangling gates, allowing for efficient use of techniques such as quantum circuit cutting. We focus on three models of quasi-1D Hamiltonians: (i) systems with impurities acting as entanglement barriers, (ii) systems with competing long-range and short-range interactions transitioning from a long-range singlet to a quantum critical state, and (iii) random quantum critical systems. For the first model, we observe a plateau in the ansatz accuracy, controlled by the number of entangling gates between subsystems. This behavior is explained by iterative capture of eigenvalues in the entanglement spectrum. In the second model, combining long-range and short-range entanglement schemes yields the best overall accuracy, leading to global convergence in the entanglement spectrum. For the third model, we use an renormalization group approach to build the short- and long-range entanglement structure of the ansatz. Our comprehensive analysis provides a new perspective on the design of ans\"atze based on the expected entanglement structure of the approximated state. | [
{
"version": "v1",
"created": "Wed, 29 Jan 2025 10:04:43 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Joch",
"Alina",
""
],
[
"Uhrig",
"Götz S.",
""
],
[
"Fauseweh",
"Benedikt",
""
]
] |
2501.17770 | Yangming Li | Yangming Li, Chaoyu Liu, Carola-Bibiane Sch\"onlieb | Generative Unordered Flow for Set-Structured Data Generation | Paper under review | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Flow-based generative models have demonstrated promising performance across a broad spectrum of data modalities (e.g., image and text). However, there are few works exploring their extension to unordered data (e.g., spatial point set), which is not trivial because previous models are mostly designed for vector data that are naturally ordered. In this paper, we present unordered flow, a type of flow-based generative model for set-structured data generation. Specifically, we convert unordered data into an appropriate function representation, and learn the probability measure of such representations through function-valued flow matching. For the inverse map from a function representation to unordered data, we propose a method similar to particle filtering, with Langevin dynamics to first warm-up the initial particles and gradient-based search to update them until convergence. We have conducted extensive experiments on multiple real-world datasets, showing that our unordered flow model is very effective in generating set-structured data and significantly outperforms previous baselines. | [
{
"version": "v1",
"created": "Wed, 29 Jan 2025 17:03:44 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:37:00 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Yangming",
""
],
[
"Liu",
"Chaoyu",
""
],
[
"Schönlieb",
"Carola-Bibiane",
""
]
] |
2501.17817 | Fran\c{c}ois Th\'eberge | Jordan Barrett, Ryan DeWolfe, Bogumi{\l} Kami\'nski, Pawe{\l} Pra{\l}at, Aaron Smith and Fran\c{c}ois Th\'eberge | Improving community detection via community association strength scores | 10 pages, 8 figures, 2 tables | null | 10.1007/978-3-031-92898-7_6 | null | cs.SI | http://creativecommons.org/licenses/by/4.0/ | Community detection methods play a central role in understanding complex networks by revealing highly connected subsets of entities. However, most community detection algorithms generate partitions of the nodes, thus (i) forcing every node to be part of a community and (ii) ignoring the possibility that some nodes may be part of multiple communities. In our work, we investigate three simple community association strength (CAS) scores and their usefulness as post-processing tools given some partition of the nodes. We show that these measures can be used to improve node partitions, detect outlier nodes (not part of any community), and help find nodes with multiple community memberships. | [
{
"version": "v1",
"created": "Wed, 29 Jan 2025 18:09:20 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Barrett",
"Jordan",
""
],
[
"DeWolfe",
"Ryan",
""
],
[
"Kamiński",
"Bogumił",
""
],
[
"Prałat",
"Paweł",
""
],
[
"Smith",
"Aaron",
""
],
[
"Théberge",
"François",
""
]
] |
2501.18032 | Fumio Abe | Fumio Abe | Fractal algorithm for multiple lens analyses | null | null | null | null | astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A microlensing exoplanet search is a unique method for finding planets orbiting distant stars. However, in the past, the method used to analyze microlensing data could not deal with complex lens systems. The number of lenses was limited three or less. Positions calculations of images and integration of them suffered from severe round-off errors because of singularities. We developed a new algorithm to calculate the light curves of multiple lens systems. In this algorithm, fractal-like consecutive self-similar division (SSD) is used to find sparse images. SSD is also useful for integrating images to efficiently obtain magnifications. The new algorithm does not use root finding for the lens equation and is free from caustic singularities. There is no limit on the number of lenses. Compared to inverse-ray shooting, this method dramatically improves the computing time. The calculation can be adjusted to obtain either a high-precision final result or high-speed quick result. Although this new algorithm was developed for a microlensing planet search, its application to quasar microlensing is also expected. This paper discusses problems in the modeling of a multiple lens system and then presents the new algorithm in detail. | [
{
"version": "v1",
"created": "Wed, 29 Jan 2025 22:39:20 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 00:21:07 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Abe",
"Fumio",
""
]
] |
2501.18282 | Yunzhen Yao | Yunzhen Yao, Lie He, Michael Gastpar | Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper considers the sample-efficiency of preference learning, which models and predicts human choices based on comparative judgments. The minimax optimal estimation error rate $\Theta(d/n)$ in classical estimation theory requires that the number of samples $n$ scales linearly with the dimensionality of the feature space $d$. However, the high dimensionality of the feature space and the high cost of collecting human-annotated data challenge the efficiency of traditional estimation methods. To remedy this, we leverage sparsity in the preference model and establish sharp error rates. We show that under the sparse random utility model, where the parameter of the reward function is $k$-sparse, the minimax optimal rate can be reduced to $\Theta(k/n \log(d/k))$. Furthermore, we analyze the $\ell_{1}$-regularized estimator and show that it achieves near-optimal rate under mild assumptions on the Gram matrix. Experiments on synthetic data and LLM alignment data validate our theoretical findings, showing that sparsity-aware methods significantly reduce sample complexity and improve prediction accuracy. | [
{
"version": "v1",
"created": "Thu, 30 Jan 2025 11:41:13 GMT"
},
{
"version": "v2",
"created": "Fri, 31 Jan 2025 02:13:55 GMT"
},
{
"version": "v3",
"created": "Mon, 26 May 2025 19:24:08 GMT"
},
{
"version": "v4",
"created": "Tue, 03 Jun 2025 20:18:22 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yao",
"Yunzhen",
""
],
[
"He",
"Lie",
""
],
[
"Gastpar",
"Michael",
""
]
] |
2501.18974 | Antonio Calcagn\`i | Antonio Calcagn\`i, Przemys{\l}aw Grzegorzewski, Maciej Romaniuk | Bayesianize Fuzziness in the Statistical Analysis of Fuzzy Data | 22 pages, 7 figures, 3 tables Note: The compressed source folder contains the Supplementary Materials | International Journal of Approximate Reasoning, Elsevier, 2025 | 10.1016/j.ijar.2025.109495 | null | stat.ME stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fuzzy data, prevalent in social sciences and other fields, capture uncertainties arising from subjective evaluations and measurement imprecision. Despite significant advancements in fuzzy statistics, a unified inferential regression-based framework remains undeveloped. Hence, we propose a novel approach for analyzing bounded fuzzy variables within a regression framework. Building on the premise that fuzzy data result from a process analogous to statistical coarsening, we introduce a conditional probabilistic approach that links observed fuzzy statistics (e.g., mode, spread) to the underlying, unobserved statistical model, which depends on external covariates. The inferential problem is addressed using Approximate Bayesian methods, mainly through a Gibbs sampler incorporating a quadratic approximation of the posterior distribution. Simulation studies and applications involving external validations are employed to evaluate the effectiveness of the proposed approach for fuzzy data analysis. By reintegrating fuzzy data analysis into a more traditional statistical framework, this work provides a significant step toward enhancing the interpretability and applicability of fuzzy statistical methods in many applicative contexts. | [
{
"version": "v1",
"created": "Fri, 31 Jan 2025 09:10:04 GMT"
},
{
"version": "v2",
"created": "Mon, 28 Apr 2025 14:46:44 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Calcagnì",
"Antonio",
""
],
[
"Grzegorzewski",
"Przemysław",
""
],
[
"Romaniuk",
"Maciej",
""
]
] |
2501.19118 | Jaksa Vucicevic | Jeremija Kova\v{c}evi\'c, Michel Ferrero and Jak\v{s}a Vu\v{c}i\v{c}evi\'c | Towards numerically exact computation of conductivity in the thermodynamic limit of interacting lattice models | 7 pages, 3 figures + supplemental material, 15 pages, 19 figures | null | null | null | cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Computing dynamical response functions in interacting lattice models is a long standing challenge in condensed matter physics. In view of recent results, the dc resistivity $\rho_\mathrm{dc}$ in the weak coupling regime of the Hubbard model is of great interest, yet it is not fully understood. The challenge lies in having to work with large lattices while avoiding analytical continuation. The weak-coupling $\rho_\mathrm{dc}$ results were so far computed at the level of the Boltzmann theory and at the level of the Kubo bubble approximation, which neglects vertex corrections. Neither theory was so far rigorously proven to give exact results even at infinitesimal coupling, and the respective dc resistivity results differ greatly. In this work we develop, cross-check and apply two state-of-the-art methods for obtaining dynamical response functions. We compute the optical conductivity at weak coupling in the Hubbard model in a fully controlled way, in the thermodynamic limit and without analytical continuation. We show that vertex corrections persist to infinitesimal coupling, with a constant ratio to the Kubo bubble. We connect our methods with the Boltzmann theory, and show that the latter applies additional approximations, which lead to quantitatively incorrect scaling of $\rho_\mathrm{dc}$ with respect to the coupling constant. | [
{
"version": "v1",
"created": "Fri, 31 Jan 2025 13:22:13 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Apr 2025 08:52:40 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 12:52:40 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kovačević",
"Jeremija",
""
],
[
"Ferrero",
"Michel",
""
],
[
"Vučičević",
"Jakša",
""
]
] |
2502.00338 | Yuan Gao | Yuan Gao, Hao Wu, Ruiqi Shu, Huanshuo Dong, Fan Xu, Rui Ray Chen, Yibo Yan, Qingsong Wen, Xuming Hu, Kun Wang, Jiahao Wu, Qing Li, Hui Xiong, Xiaomeng Huang | OneForecast: A Universal Framework for Global and Regional Weather Forecasting | null | null | null | null | cs.LG physics.ao-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate weather forecasts are important for disaster prevention, agricultural planning, etc. Traditional numerical weather prediction (NWP) methods offer physically interpretable high-accuracy predictions but are computationally expensive and fail to fully leverage rapidly growing historical data. In recent years, deep learning models have made significant progress in weather forecasting, but challenges remain, such as balancing global and regional high-resolution forecasts, excessive smoothing in extreme event predictions, and insufficient dynamic system modeling. To address these issues, this paper proposes a global-regional nested weather forecasting framework (OneForecast) based on graph neural networks. By combining a dynamic system perspective with multi-grid theory, we construct a multi-scale graph structure and densify the target region to capture local high-frequency features. We introduce an adaptive messaging mechanism, using dynamic gating units to deeply integrate node and edge features for more accurate extreme event forecasting. For high-resolution regional forecasts, we propose a neural nested grid method to mitigate boundary information loss. Experimental results show that OneForecast performs excellently across global to regional scales and short-term to long-term forecasts, especially in extreme event predictions. Codes link https://github.com/YuanGao-YG/OneForecast. | [
{
"version": "v1",
"created": "Sat, 01 Feb 2025 06:49:16 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 05:33:10 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gao",
"Yuan",
""
],
[
"Wu",
"Hao",
""
],
[
"Shu",
"Ruiqi",
""
],
[
"Dong",
"Huanshuo",
""
],
[
"Xu",
"Fan",
""
],
[
"Chen",
"Rui Ray",
""
],
[
"Yan",
"Yibo",
""
],
[
"Wen",
"Qingsong",
""
],
[
"Hu",
"Xuming",
""
],
[
"Wang",
"Kun",
""
],
[
"Wu",
"Jiahao",
""
],
[
"Li",
"Qing",
""
],
[
"Xiong",
"Hui",
""
],
[
"Huang",
"Xiaomeng",
""
]
] |
2502.00471 | Pavel Chebotarev | Pavel Chebotarev | Evolution of Society Caused by Collective and Individual Decisions | 15 pages, 9 figures, a converence paper. Accepted for Springer Lecture Notes in Business Information Processing | In: Moreno-Jim\'enez, J.M., Costa Morais, D., Escobar, M.T., Tur\'on, A. (eds) Human and Artificial Intelligence in Group Decision and Negotiation. GDN 2025. Lecture Notes in Business Information Processing, vol 553. Springer, Cham | 10.1007/978-3-031-95221-0_3 | null | physics.soc-ph cs.GT math.OC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Decision-making societies may vary in their level of cooperation and degree of conservatism, both of which influence their overall performance. Moreover, these factors are not fixed -- they can change based on the decisions agents in the society make in their interests. But can these changes lead to cyclical patterns in societal evolution? To explore this question, we use the ViSE (Voting in Stochastic Environment) model. In this framework, the level of cooperation can be measured by group size, while the degree of conservatism is determined by the voting threshold. Agents can adopt either individualistic or group-oriented strategies when voting on stochastically generated external proposals. For Gaussian proposal generators, the expected capital gain (ECG) -- a measure of agents' performance -- can be expressed in standard mathematical functions. Our findings show that in neutral environments, societal evolution with open or democratic groups can follow cyclic patterns. We also find that highly conservative societies or conservative societies with low levels of cooperation can evolve into liberal (less conservative than majoritarian) societies and that mafia groups never let their members go when they want to. | [
{
"version": "v1",
"created": "Sat, 01 Feb 2025 15:56:19 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Mar 2025 11:31:20 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Apr 2025 07:47:46 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chebotarev",
"Pavel",
""
]
] |
2502.00514 | Shuyang Gong | Hang Du, Shuyang Gong, Jiaming Xu | A Proof of The Changepoint Detection Threshold Conjecture in Preferential Attachment Models | Added more discussion on background and proof ideas; conference version accepted by COLT 2025 | null | null | null | math.PR math.CO math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the problem of detecting and estimating a changepoint in the attachment function of a network evolving according to a preferential attachment model on $n$ vertices, using only a single final snapshot of the network. Bet et al.~\cite{bet2023detecting} show that a simple test based on thresholding the number of vertices with minimum degrees can detect the changepoint when the change occurs at time $n-\Omega(\sqrt{n})$. They further make the striking conjecture that detection becomes impossible for any test if the change occurs at time $n-o(\sqrt{n}).$ Kaddouri et al.~\cite{kaddouri2024impossibility} make a step forward by proving the detection is impossible if the change occurs at time $n-o(n^{1/3}).$ In this paper, we resolve the conjecture affirmatively, proving that detection is indeed impossible if the change occurs at time $n-o(\sqrt{n}).$ Furthermore, we establish that estimating the changepoint with an error smaller than $o(\sqrt{n})$ is also impossible, thereby confirming that the estimator proposed in Bhamidi et al.~\cite{bhamidi2018change} is order-optimal. | [
{
"version": "v1",
"created": "Sat, 01 Feb 2025 18:19:36 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:27:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Du",
"Hang",
""
],
[
"Gong",
"Shuyang",
""
],
[
"Xu",
"Jiaming",
""
]
] |
2502.00675 | Minghang Deng | Minghang Deng, Ashwin Ramachandran, Canwen Xu, Lanxiang Hu, Zhewei Yao, Anupam Datta, Hao Zhang | ReFoRCE: A Text-to-SQL Agent with Self-Refinement, Consensus Enforcement, and Column Exploration | 33 pages, 3 figures | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present ReFoRCE, a Text-to-SQL agent that tops the Spider 2.0 leaderboard--a challenging benchmark reflecting complex, real-world Text-to-SQL scenarios. While Text-to-SQL systems enable natural language queries over structured databases, deploying them in enterprise environments remains difficult due to large, complex schemas (with over 1,000 columns), diverse SQL dialects (e.g., BigQuery, Snowflake), and sophisticated query requirements (e.g., transformations and analytics). ReFoRCE addresses these challenges through: (a) database information compression via pattern-based table grouping and LLM-guided schema linking to alleviate long-context issues; (b) self-refinement to iteratively correct syntax and semantic errors across dialects; (c) majority-vote consensus to select high-confidence candidates while deferring ambiguous cases arising from sophisticated queries; and (d) iterative column exploration guided by execution feedback to resolve those deferred cases. ReFoRCE achieves new state-of-the-art results, with scores of 35.83 on Spider 2.0-Snow and 36.56 on Spider 2.0-Lite. | [
{
"version": "v1",
"created": "Sun, 02 Feb 2025 05:25:03 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Feb 2025 08:38:16 GMT"
},
{
"version": "v3",
"created": "Fri, 11 Apr 2025 22:42:10 GMT"
},
{
"version": "v4",
"created": "Thu, 22 May 2025 05:41:29 GMT"
},
{
"version": "v5",
"created": "Tue, 03 Jun 2025 18:03:32 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Deng",
"Minghang",
""
],
[
"Ramachandran",
"Ashwin",
""
],
[
"Xu",
"Canwen",
""
],
[
"Hu",
"Lanxiang",
""
],
[
"Yao",
"Zhewei",
""
],
[
"Datta",
"Anupam",
""
],
[
"Zhang",
"Hao",
""
]
] |
2502.00698 | Huanqia Cai | Huanqia Cai, Yijun Yang, Winston Hu | MM-IQ: Benchmarking Human-Like Abstraction and Reasoning in Multimodal Models | null | null | null | null | cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | IQ testing has served as a foundational methodology for evaluating human cognitive capabilities, deliberately decoupling assessment from linguistic background, language proficiency, or domain-specific knowledge to isolate core competencies in abstraction and reasoning. Yet, artificial intelligence research currently lacks systematic benchmarks to quantify these critical cognitive capabilities in multimodal systems. To address this crucial gap, we propose MM-IQ, a comprehensive evaluation framework, which comprises a large-scale training set with 4,776 visual reasoning problems and 2,710 meticulously curated test items spanning 8 distinct reasoning paradigms. Through systematic evaluation of existing open-source and proprietary multimodal models, our benchmark reveals striking limitations: even state-of-the-art architectures achieve only marginally superior performance to random chance (33.17% vs. 25% baseline accuracy). This substantial performance chasm highlights the inadequacy of current multimodal models in approximating fundamental human reasoning capacities, underscoring the need for paradigm-shifting advancements to bridge this cognitive divide. Moreover, inspired by the recent surge of large reasoning models, we also release a multimodal reasoning model as the baseline that is trained via reinforcement learning with verifiable reward functions, reaching competitive performance to the state-of-the-art with a notably smaller model size. | [
{
"version": "v1",
"created": "Sun, 02 Feb 2025 07:12:03 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:20:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cai",
"Huanqia",
""
],
[
"Yang",
"Yijun",
""
],
[
"Hu",
"Winston",
""
]
] |
2502.01203 | Gholamali Aminian | Gholamali Aminian, Amir R. Asadi, Idan Shenfeld, Youssef Mroueh | Theoretical Analysis of KL-regularized RLHF with Multiple Reference Models | Experiments are added in new version | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Recent methods for aligning large language models (LLMs) with human feedback predominantly rely on a single reference model, which limits diversity, model overfitting, and underutilizes the wide range of available pre-trained models. Incorporating multiple reference models has the potential to address these limitations by broadening perspectives, reducing bias, and leveraging the strengths of diverse open-source LLMs. However, integrating multiple reference models into reinforcement learning with human feedback (RLHF) frameworks poses significant theoretical challenges, where achieving exact solutions has remained an open problem. This paper presents the first \emph{exact solution} to the multiple reference model problem in reverse KL-regularized RLHF. We introduce a comprehensive theoretical framework that includes rigorous statistical analysis and provides sample complexity guarantees. Additionally, we extend our analysis to forward KL-regularized RLHF, offering new insights into sample complexity requirements in multiple reference scenarios. Our contributions lay the foundation for more advanced and adaptable LLM alignment techniques, enabling the effective use of multiple reference models. This work paves the way for developing alignment frameworks that are both theoretically sound and better suited to the challenges of modern AI ecosystems. | [
{
"version": "v1",
"created": "Mon, 03 Feb 2025 09:50:30 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 13:43:34 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Aminian",
"Gholamali",
""
],
[
"Asadi",
"Amir R.",
""
],
[
"Shenfeld",
"Idan",
""
],
[
"Mroueh",
"Youssef",
""
]
] |
2502.01419 | Mingi Jung | Mingi Jung, Saehyung Lee, Eunji Kim, and Sungroh Yoon | Visual Attention Never Fades: Selective Progressive Attention ReCalibration for Detailed Image Captioning in Multimodal Large Language Models | ICML 2025 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Detailed image captioning is essential for tasks like data generation and aiding visually impaired individuals. High-quality captions require a balance between precision and recall, which remains challenging for current multimodal large language models (MLLMs). In this work, we hypothesize that this limitation stems from weakening and increasingly noisy visual attention as responses lengthen. To address this issue, we propose SPARC (Selective Progressive Attention ReCalibration), a training-free method that enhances the contribution of visual tokens during decoding. SPARC is founded on three key observations: (1) increasing the influence of all visual tokens reduces recall; thus, SPARC selectively amplifies visual tokens; (2) as captions lengthen, visual attention becomes noisier, so SPARC identifies critical visual tokens by leveraging attention differences across time steps; (3) as visual attention gradually weakens, SPARC reinforces it to preserve its influence. Our experiments, incorporating both automated and human evaluations, demonstrate that existing methods improve the precision of MLLMs at the cost of recall. In contrast, our proposed method enhances both precision and recall with minimal computational overhead. | [
{
"version": "v1",
"created": "Mon, 03 Feb 2025 14:58:11 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 06:43:40 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jung",
"Mingi",
""
],
[
"Lee",
"Saehyung",
""
],
[
"Kim",
"Eunji",
""
],
[
"Yoon",
"Sungroh",
""
]
] |
2502.01454 | Andr\'e Oct\'avio Soares | Andr\'e O. Soares, Anna Ritz-Zwilling, Jean-No\"el Fuchs | Extended string-net models with all anyons at finite temperature | 19 pages, 7 figures, 1 Table | Phys. Rev. B 111, 235405 (2025) | 10.1103/PhysRevB.111.235405 | null | cond-mat.mes-hall cond-mat.str-el quant-ph | http://creativecommons.org/licenses/by/4.0/ | The string-net model describes a vast family of topological orders in two spatial dimensions. Here, we consider the effect of thermal fluctuations on these topological phases. In the original string-net model, the description of charge (vertex) excitations can be problematic. Therefore, in order to describe all anyon excitations, we study an extended model [Y. Hu et al., Phys. Rev. B 97, 195154 (2018)]. Building on recent methods, we compute the spectral degeneracies of excited states and obtain the exact partition function. In the thermodynamic limit, the latter is dominated by the trivial (vacuum) anyon, so that topological order is destroyed at any nonzero temperature. In contrast, in a finite-size system, order survives up to a finite temperature, revealing a nontrivial scaling between temperature and size similar to that of the one-dimensional classical Ising model. We confirm this scaling by computing the thermal average of several observables such as Wegner-Wilson loops and topological mutual information. | [
{
"version": "v1",
"created": "Mon, 03 Feb 2025 15:43:19 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 17:17:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Soares",
"André O.",
""
],
[
"Ritz-Zwilling",
"Anna",
""
],
[
"Fuchs",
"Jean-Noël",
""
]
] |
2502.01458 | Wei Yao | Wei Yao, Wenkai Yang, Gengze Xu, Ziqiao Wang, Yankai Lin, Yong Liu | The Capabilities and Limitations of Weak-to-Strong Generalization: Generalization and Calibration | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Weak-to-strong generalization, where weakly supervised strong models outperform their weaker teachers, offers a promising approach to aligning superhuman models with human values. To deepen the understanding of this approach, we provide theoretical insights into its capabilities and limitations. First, in the classification setting, we establish upper and lower generalization error bounds for the strong model, identifying the primary limitations as stemming from the weak model's generalization error and the optimization objective itself. Additionally, we derive lower and upper bounds on the calibration error of the strong model. These theoretical bounds reveal two critical insights: (1) the weak model should demonstrate strong generalization performance and maintain well-calibrated predictions, and (2) the strong model's training process must strike a careful balance, as excessive optimization could undermine its generalization capability by over-relying on the weak supervision signals. Finally, in the regression setting, we extend the work of Charikar et al. (2024) to a loss function based on Kullback-Leibler (KL) divergence, offering guarantees that the strong student can outperform its weak teacher by at least the magnitude of their disagreement. We conduct sufficient experiments to validate our theory. | [
{
"version": "v1",
"created": "Mon, 03 Feb 2025 15:48:28 GMT"
},
{
"version": "v2",
"created": "Sat, 01 Mar 2025 08:45:16 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 01:52:51 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yao",
"Wei",
""
],
[
"Yang",
"Wenkai",
""
],
[
"Xu",
"Gengze",
""
],
[
"Wang",
"Ziqiao",
""
],
[
"Lin",
"Yankai",
""
],
[
"Liu",
"Yong",
""
]
] |
2502.03535 | Mohammadhossein Dadgar | Mohammadhossein Dadgar, Christopher L. Baldwin | The anomalously slow dynamics of inhomogeneous quantum annealing | null | Phys. Rev. A 111, 062601 (2025) | 10.1103/PhysRevA.111.062601 | null | quant-ph cond-mat.dis-nn cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Inhomogeneous quantum annealing (IQA), in which transverse fields are turned off one by one rather than simultaneously, has been proposed as an effective way to avoid the first-order phase transitions that impede conventional quantum annealing (QA). Here we explicitly study the dynamics of IQA, rather than merely the thermodynamics, and find that it is appreciably slower than the phase diagram would suggest. Interestingly, this slowdown manifests both when IQA succeeds in circumventing phase transitions and when it fails. Even in the absence of transitions, such as for the mean-field models that have been analyzed previously, IQA is slower than expected by a factor of the number of spins $N$. More significantly, we show that in non-mean-field models, first-order transitions are likely to be quite common, and the gap at such transitions is not merely exponential in $N$ but exactly zero. Thus IQA cannot reach the ground state on any timescale. Both of these results can be understood through the simple observation that a spin's magnetization becomes conserved once its field is turned off during the IQA protocol. | [
{
"version": "v1",
"created": "Wed, 05 Feb 2025 19:00:04 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:00:02 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dadgar",
"Mohammadhossein",
""
],
[
"Baldwin",
"Christopher L.",
""
]
] |
2502.03783 | Luohong Wu | Luohong Wu, Nicola A. Cavalcanti, Matthias Seibold, Giuseppe Loggia, Lisa Reissner, Jonas Hein, Silvan Beeler, Arnd Vieh\"ofer, Stephan Wirth, Lilian Calvet, Philipp F\"urnstahl | UltraBones100k: A reliable automated labeling method and large-scale dataset for ultrasound-based bone surface extraction | accepted by Computers in Biology and Medicine | Computers in Biology and Medicine 194 (2025) 110435 | 10.1016/j.compbiomed.2025.110435. | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Ultrasound-based bone surface segmentation is crucial in computer-assisted orthopedic surgery. However, ultrasound images have limitations, including a low signal-to-noise ratio, and acoustic shadowing, which make interpretation difficult. Existing deep learning models for bone segmentation rely primarily on costly manual labeling by experts, limiting dataset size and model generalizability. Additionally, the complexity of ultrasound physics and acoustic shadow makes the images difficult for humans to interpret, leading to incomplete labels in anechoic regions and limiting model performance. To advance ultrasound bone segmentation and establish effective model benchmarks, larger and higher-quality datasets are needed.
We propose a methodology for collecting ex-vivo ultrasound datasets with automatically generated bone labels, including anechoic regions. The proposed labels are derived by accurately superimposing tracked bone CT models onto the tracked ultrasound images. These initial labels are refined to account for ultrasound physics. A clinical evaluation is conducted by an expert physician specialized on orthopedic sonography to assess the quality of the generated bone labels. A neural network for bone segmentation is trained on the collected dataset and its predictions are compared to expert manual labels, evaluating accuracy, completeness, and F1-score.
We collected the largest known dataset of 100k ultrasound images of human lower limbs with bone labels, called UltraBones100k. A Wilcoxon signed-rank test with Bonferroni correction confirmed that the bone alignment after our method significantly improved the quality of bone labeling (p < 0.001). The model trained on UltraBones100k consistently outperforms manual labeling in all metrics, particularly in low-intensity regions (320% improvement in completeness at a distance threshold of 0.5 mm). | [
{
"version": "v1",
"created": "Thu, 06 Feb 2025 05:11:29 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Feb 2025 07:03:17 GMT"
},
{
"version": "v3",
"created": "Tue, 27 May 2025 10:44:53 GMT"
},
{
"version": "v4",
"created": "Wed, 04 Jun 2025 14:49:21 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wu",
"Luohong",
""
],
[
"Cavalcanti",
"Nicola A.",
""
],
[
"Seibold",
"Matthias",
""
],
[
"Loggia",
"Giuseppe",
""
],
[
"Reissner",
"Lisa",
""
],
[
"Hein",
"Jonas",
""
],
[
"Beeler",
"Silvan",
""
],
[
"Viehöfer",
"Arnd",
""
],
[
"Wirth",
"Stephan",
""
],
[
"Calvet",
"Lilian",
""
],
[
"Fürnstahl",
"Philipp",
""
]
] |
2502.03956 | Francesco Pontiggia | Francesco Pontiggia, Ezio Bartocci, Michele Chiari | POPACheck: A Model Checker for Probabilistic Pushdown Automata | 16 pages, 10 Figures, 4 Tables. Accepted for publication in the Proceedings of the 37th International Conference on Computer Aided Verification (CAV'25) | null | null | null | cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present POPACheck, the first model checking tool for probabilistic Pushdown Automata (pPDA) supporting temporal logic specifications. POPACheck provides a user-friendly probabilistic modeling language with recursion that automatically translates into Probabilistic Operator Precedence Automata (pOPA). pOPA are a class of pPDA that can express all the behaviors of probabilistic programs: sampling, conditioning, recursive procedures, and nested inference queries. On pOPA, POPACheck can solve reachability queries as well as qualitative and quantitative model checking queries for specifications in Linear Temporal Logic (LTL) and a fragment of Precedence Oriented Temporal Logic (POTL), a logic for context-free properties such as pre/post-conditioning. | [
{
"version": "v1",
"created": "Thu, 06 Feb 2025 10:47:26 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:20:54 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Pontiggia",
"Francesco",
""
],
[
"Bartocci",
"Ezio",
""
],
[
"Chiari",
"Michele",
""
]
] |
2502.05918 | Seok Hyun Byun | Seok Hyun Byun, Wayne Goddard | A Note on One-Hole Domino Tilings of Squares and Rectangles | 12 pages, 6 figures, some typos are fixed | null | null | null | math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the number of domino tilings of an odd-by-odd rectangle that leave one hole. This problem is equivalent to the number of near-perfect matchings of the odd-by-odd rectangular grid. For any particular position of the vacancy on the $(2k+1)\times (2k+1)$ square grid, we show that the number of near-perfect matchings is a multiple of $2^k$, and from this follows a conjecture of Kong that the total number of near-perfect matchings is a multiple of $2^k$. We also determine the parity of the number of near-perfect matchings with a particular vacancy for the rectangle case. | [
{
"version": "v1",
"created": "Sun, 09 Feb 2025 14:26:42 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 20:07:00 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Byun",
"Seok Hyun",
""
],
[
"Goddard",
"Wayne",
""
]
] |
2502.06244 | Zeman Li | Zeman Li, Yuan Deng, Peilin Zhong, Meisam Razaviyayn, Vahab Mirrokni | PiKE: Adaptive Data Mixing for Large-Scale Multi-Task Learning Under Low Gradient Conflicts | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Modern foundation models are trained on diverse datasets to enhance generalization across tasks and domains A central challenge in this process is determining how to effectively mix and sample data from multiple sources This naturally leads to a multitask learning (MTL) perspective While prior work in MTL has emphasized mitigating gradient conflicts we observe that largescale pretraining scenariossuch as multilingual or multidomain trainingoften exhibit little to no gradient conflict Motivated by this observation we propose PiKE (Positive gradient interaction-based K-task weights Estimator) an adaptive data mixing algorithm that dynamically adjusts sampling weights during training PiKE exploits nonconflicting gradient interactions to minimize a neartight upper bound on the average loss decrease at each step while incurring negligible computational overhead We provide theoretical convergence guarantees and show that PiKE outperforms static and nonadaptive mixing baselines Furthermore we extend PiKE to promote balanced learning across tasks Extensive experiments on largescale language model pretraining confirm that PiKE achieves faster convergence and improved downstream performance compared to existing approaches | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 08:23:05 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 02:16:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Zeman",
""
],
[
"Deng",
"Yuan",
""
],
[
"Zhong",
"Peilin",
""
],
[
"Razaviyayn",
"Meisam",
""
],
[
"Mirrokni",
"Vahab",
""
]
] |
2502.06597 | Nikita P. Kalinin | Nikita P. Kalinin, Jalaj Upadhyay, Christoph H. Lampert | Continual Release Moment Estimation with Differential Privacy | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose Joint Moment Estimation (JME), a method for continually and privately estimating both the first and second moments of data with reduced noise compared to naive approaches. JME uses the matrix mechanism and a joint sensitivity analysis to allow the second moment estimation with no additional privacy cost, thereby improving accuracy while maintaining privacy. We demonstrate JME's effectiveness in two applications: estimating the running mean and covariance matrix for Gaussian density estimation, and model training with DP-Adam on CIFAR-10. | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 15:58:26 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:40:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kalinin",
"Nikita P.",
""
],
[
"Upadhyay",
"Jalaj",
""
],
[
"Lampert",
"Christoph H.",
""
]
] |
2502.07045 | John Hastings | Haywood Gelman, John D. Hastings | Scalable and Ethical Insider Threat Detection through Data Synthesis and Analysis by LLMs | 6 pages, 0 figures, 8 tables | 2025 IEEE 13th International Symposium on Digital Forensics and Security (ISDFS) | 10.1109/ISDFS65363.2025.11012066 | null | cs.CR cs.AI cs.CL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Insider threats wield an outsized influence on organizations, disproportionate to their small numbers. This is due to the internal access insiders have to systems, information, and infrastructure. %One example of this influence is where anonymous respondents submit web-based job search site reviews, an insider threat risk to organizations. Signals for such risks may be found in anonymous submissions to public web-based job search site reviews. This research studies the potential for large language models (LLMs) to analyze and detect insider threat sentiment within job site reviews. Addressing ethical data collection concerns, this research utilizes synthetic data generation using LLMs alongside existing job review datasets. A comparative analysis of sentiment scores generated by LLMs is benchmarked against expert human scoring. Findings reveal that LLMs demonstrate alignment with human evaluations in most cases, thus effectively identifying nuanced indicators of threat sentiment. The performance is lower on human-generated data than synthetic data, suggesting areas for improvement in evaluating real-world data. Text diversity analysis found differences between human-generated and LLM-generated datasets, with synthetic data exhibiting somewhat lower diversity. Overall, the results demonstrate the applicability of LLMs to insider threat detection, and a scalable solution for insider sentiment testing by overcoming ethical and logistical barriers tied to data acquisition. | [
{
"version": "v1",
"created": "Mon, 10 Feb 2025 21:27:06 GMT"
},
{
"version": "v2",
"created": "Mon, 07 Apr 2025 16:01:47 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gelman",
"Haywood",
""
],
[
"Hastings",
"John D.",
""
]
] |
2502.07185 | Prakhar Bansal | Prakhar Bansal and Dragan Huterer | Expansion-history preferences of DESI DR2 and external data | 5 pages, 2 figures, plus appendices (updated with DESI DR2 data; results qualitatively unchanged) | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We explore the origin of the preference of DESI DR2 baryon acoustic oscillation (BAO) measurements and external data from cosmic microwave background (CMB) and type Ia supernovae (SNIa) that dark energy behavior departs from that expected in the standard cosmological model with vacuum energy ($\Lambda$CDM). In our analysis, we allow a flexible scaling of the expansion rate with redshift that nevertheless allows reasonably tight constraints on the quantities of interest, and adopt and validate a simple yet accurate compression of the CMB data that allows us to constrain our phenomenological model of the expansion history. We find that data consistently show a preference for a 3-4 % increase in the expansion rate at $z\simeq 0.7$ relative to that predicted by the standard $\Lambda$CDM model, in excellent agreement with results from the less flexible $(w_0, w_a)$ parameterization which was used in previous analyses. Even though our model allows a departure from the best-fit $\Lambda$CDM model at zero redshift, we find no evidence for such a signal. We also find no evidence (at greater than 1$\sigma$ significance) for a departure of the expansion rate from the $\Lambda$CDM predictions at higher redshifts for any of the data combinations that we consider. Overall, our results strengthen the robustness of the findings using the combination of DESI, CMB, and SNIa data to dark-energy modeling assumptions. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 02:25:40 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 21:29:30 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bansal",
"Prakhar",
""
],
[
"Huterer",
"Dragan",
""
]
] |
2502.07782 | Nathan Mankovich | Nathan Mankovich, Ignacio Santamaria, Gustau Camps-Valls, Tolga Birdal | A Flag Decomposition for Hierarchical Datasets | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Flag manifolds encode nested sequences of subspaces and serve as powerful structures for various computer vision and machine learning applications. Despite their utility in tasks such as dimensionality reduction, motion averaging, and subspace clustering, current applications are often restricted to extracting flags using common matrix decomposition methods like the singular value decomposition. Here, we address the need for a general algorithm to factorize and work with hierarchical datasets. In particular, we propose a novel, flag-based method that decomposes arbitrary hierarchical real-valued data into a hierarchy-preserving flag representation in Stiefel coordinates. Our work harnesses the potential of flag manifolds in applications including denoising, clustering, and few-shot learning. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 18:59:52 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:55:15 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mankovich",
"Nathan",
""
],
[
"Santamaria",
"Ignacio",
""
],
[
"Camps-Valls",
"Gustau",
""
],
[
"Birdal",
"Tolga",
""
]
] |
2502.07955 | Arthur Carlton-Jones | Arthur Carlton-Jones, Alonso Suarez, Yun-Suk Eo, Ian M. Hayes, Shanta R. Saha, Johnpierre Paglione, Nicholas P. Butch, and Steven M. Anlage | Revealing isotropic abundant low-energy excitations in UTe$_2$ through complex microwave surface impedance | 29 pages, 23 figures | null | null | null | cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | The complex surface impedance is a well-established tool to study the super- and normal-fluid responses of superconductors. Fundamental properties of the superconductor, such as the pairing mechanism, Fermi surface, and topological properties, also influence the surface impedance. We explore the microwave surface impedance of spin-triplet UTe$_2$ single crystals as a function of temperature using resonant cavity perturbation measurements employing a novel multi-modal analysis to gain insight into these properties. We determine a composite surface impedance of the crystal for each mode using resonance data combined with the independently measured normal state dc resistivity tensor. The normal state surface impedance reveals the weighting of current flow directions in the crystal of each resonant mode. For UTe$_2$, we find an isotropic $\Delta \lambda(T) \sim T^\alpha$ power-law temperature dependence for the magnetic penetration depth for $T\le T_c/3$ with $\alpha < 2$, which is inconsistent with a single pair of point nodes on the Fermi surface under weak scattering. We also find a similar power-law temperature dependence for the low-temperature surface resistance $R_s(T) \sim T^{\alpha_R}$ with $\alpha_R < 2$. We observe a strong anisotropy of the residual microwave loss across these modes, with some modes showing loss below the universal line-nodal value, to those showing substantially more. We compare to predictions for topological Weyl superconductivity in the context of the observed isotropic power-laws, and anisotropy of the residual loss. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 21:06:58 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 06:30:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Carlton-Jones",
"Arthur",
""
],
[
"Suarez",
"Alonso",
""
],
[
"Eo",
"Yun-Suk",
""
],
[
"Hayes",
"Ian M.",
""
],
[
"Saha",
"Shanta R.",
""
],
[
"Paglione",
"Johnpierre",
""
],
[
"Butch",
"Nicholas P.",
""
],
[
"Anlage",
"Steven M.",
""
]
] |
2502.07975 | Oliver Biggar | Oliver Biggar and Christos Papadimitriou | Sink equilibria and the attractors of learning in games | null | null | null | null | cs.GT cs.LG | http://creativecommons.org/licenses/by/4.0/ | Characterizing the limit behavior -- that is, the attractors -- of learning dynamics is one of the most fundamental open questions in game theory. In recent work in this front, it was conjectured that the attractors of the replicator dynamic are in one-to-one correspondence with the sink equilibria of the game -- the sink strongly connected components of a game's preference graph -- , and it was established that they do stand in at least one-to-many correspondence with them. We make threefold progress on the problem of characterizing attractors. First, we show through a topological construction that the one-to-one conjecture is false. The counterexamples derive from objects called local sources -- fixed points which lie within the sink equilibrium yet are locally repelling. Second, we make progress on the attractor characterization problem for two-player games by establishing that the one-to-one conjecture is true when a local property called pseudoconvexity holds. Pseudoconvexity prevents the existence of local sources, and generalizes the existing cases -- such as zero-sum games and potential games -- where the conjecture was known to be true. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 21:40:11 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 20:40:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Biggar",
"Oliver",
""
],
[
"Papadimitriou",
"Christos",
""
]
] |
2502.08001 | Haonan Shi | Haonan Shi, Tu Ouyang, An Wang | Unveiling Client Privacy Leakage from Public Dataset Usage in Federated Distillation | To appear in Proceedings of Privacy Enhancing Technologies 2025 | null | null | null | cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated Distillation (FD) has emerged as a popular federated training framework, enabling clients to collaboratively train models without sharing private data. Public Dataset-Assisted Federated Distillation (PDA-FD), which leverages public datasets for knowledge sharing, has become widely adopted. Although PDA-FD enhances privacy compared to traditional Federated Learning, we demonstrate that the use of public datasets still poses significant privacy risks to clients' private training data. This paper presents the first comprehensive privacy analysis of PDA-FD in presence of an honest-but-curious server. We show that the server can exploit clients' inference results on public datasets to extract two critical types of private information: label distributions and membership information of the private training dataset. To quantify these vulnerabilities, we introduce two novel attacks specifically designed for the PDA-FD setting: a label distribution inference attack and innovative membership inference methods based on Likelihood Ratio Attack (LiRA). Through extensive evaluation of three representative PDA-FD frameworks (FedMD, DS-FL, and Cronus), our attacks achieve state-of-the-art performance, with label distribution attacks reaching minimal KL-divergence and membership inference attacks maintaining high True Positive Rates under low False Positive Rate constraints. Our findings reveal significant privacy risks in current PDA-FD frameworks and emphasize the need for more robust privacy protection mechanisms in collaborative learning systems. | [
{
"version": "v1",
"created": "Tue, 11 Feb 2025 22:48:49 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 15:19:15 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Shi",
"Haonan",
""
],
[
"Ouyang",
"Tu",
""
],
[
"Wang",
"An",
""
]
] |
2502.08072 | Xiang Liu | Zi-Long Man, Si-Qiang Luo, Zi-Yue Bai and Xiang Liu | Coupled-channel study of $4S$-$3D$ mixing dynamics in $\psi(4220)$ and $\psi(4380)$ | 12 pages, 4 figures and 3 tables. More discussions added. Accepted by Phys. Lett. B | null | null | null | hep-ph hep-ex | http://creativecommons.org/publicdomain/zero/1.0/ | Among charmoniumlike $XYZ$ states, the $\psi(4220)$ and $\psi(4380)$ states have emerged as key candidates for exploring the charmonium spectrum. In this work, we propose a $4S$-$3D$ charmonium mixing scheme for the $\psi(4220)$ and $\psi(4380)$, induced by coupled-channel effects. By constructing a coupled-channel model, we identify the dynamical mechanism responsible for the large mixing angle observed in previous studies, which cannot be explained by conventional potential models alone. Our analysis reveals that the $DD_1$ channel significantly influences the lower state ($\psi(4220)$), while the $D^*D_1$ channel primarily affects the higher state ($\psi(4380)$). Furthermore, we investigate the two-body Okubo-Zweig-Iizuka (OZI)-allowed strong decay behaviors of these states, providing insights into their total widths. This study not only supports the $4S$-$3D$ mixing scheme but also offers a deeper understanding of the role of coupled channels in shaping the charmonium spectrum above 4 GeV. Our results align with experimental observations and provide a framework for interpreting future data on charmonium states. | [
{
"version": "v1",
"created": "Wed, 12 Feb 2025 02:26:45 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 08:45:07 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Man",
"Zi-Long",
""
],
[
"Luo",
"Si-Qiang",
""
],
[
"Bai",
"Zi-Yue",
""
],
[
"Liu",
"Xiang",
""
]
] |
2502.08660 | Huiyao Chen | Huiyao Chen, Meishan Zhang, Jing Li, Min Zhang, Lilja {\O}vrelid, Jan Haji\v{c}, Hao Fei | Semantic Role Labeling: A Systematical Survey | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Semantic role labeling (SRL) is a central natural language processing (NLP) task aiming to understand the semantic roles within texts, facilitating a wide range of downstream applications. While SRL has garnered extensive and enduring research, there is currently a lack of a comprehensive survey that thoroughly organizes and synthesizes the field. This paper aims to review the entire research trajectory of the SRL community over the past two decades. We begin by providing a complete definition of SRL. To offer a comprehensive taxonomy, we categorize SRL methodologies into four key perspectives: model architectures, syntax feature modeling, application scenarios, and multi-modal extensions. Further, we discuss SRL benchmarks, evaluation metrics, and paradigm modeling approaches, while also exploring practical applications across various domains. Finally, we analyze future research directions in SRL, addressing the evolving role of SRL in the age of large language models (LLMs) and its potential impact on the broader NLP landscape. We maintain a public repository and consistently update related resources at: https://github.com/DreamH1gh/Awesome-SRL | [
{
"version": "v1",
"created": "Sun, 09 Feb 2025 12:45:49 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Feb 2025 06:32:15 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chen",
"Huiyao",
""
],
[
"Zhang",
"Meishan",
""
],
[
"Li",
"Jing",
""
],
[
"Zhang",
"Min",
""
],
[
"Øvrelid",
"Lilja",
""
],
[
"Hajič",
"Jan",
""
],
[
"Fei",
"Hao",
""
]
] |
2502.08748 | Ali Kaya | Ali Kaya | Vacuum Polarization, Geodesic Equation and Sachs-Wolfe Effect | v2: 6 pages, revtex4; references, comments and a figure added | null | null | null | astro-ph.CO gr-qc hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that the null geodesic equation for photons is modified in the presence of a charged scalar field, with quantum fluctuations acting as an effective mass term that changes the null paths to timelike curves. This effect can be interpreted as a vacuum polarization phenomenon in curved spacetime. The resulting contribution to the Sachs-Wolfe effect varies with photon frequency, leading to frequency-dependent corrections to the cosmic microwave background (CMB) blackbody spectrum in the form of a $\mu$-distortion, as well as modifications to the CMB power spectrum. We estimate these within a standard inflationary scenario and find that while the correction to the CMB power spectrum is significant when the scalar field is light, the magnitude of the $\mu$-distortion depends strongly on the regularization prescription. | [
{
"version": "v1",
"created": "Wed, 12 Feb 2025 19:38:21 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 20:30:09 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kaya",
"Ali",
""
]
] |
2502.08905 | Tangyu Jiang | Tangyu Jiang, Haodi Wang, Chun Yuan | DiffoRA: Enabling Parameter-Efficient Fine-Tuning via Differential Module Selection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The Parameter-Efficient Fine-Tuning (PEFT) methods have been extensively researched for large language models in downstream tasks. Among all the existing approaches, the Low-Rank Adaptation (LoRA) has gained popularity for its streamlined design by incorporating low-rank matrices into existing pre-trained models. Though effective, LoRA, as well as its adaptive optimizations, either allocate the same matrix to all the modules or adjust the interior rank of the components based on importance scoring indicators. In this paper, we argue that not all the modules in LLMs are suitable and necessary to be fine-tuned. Enlightened by this insight, we propose a new PEFT scheme called DiffoRA, which enables adaptive adoption of the low-rank decomposition matrices. At the core of DiffoRA lies a Differential Adaptation Matrix (DAM) to determine which module is the most suitable and essential for fine-tuning. We theoretically explain how the designed matrix impacts the convergence rate and generalization capability of a pre-trained model. We then construct the DAM via continuous relaxation and discretization with weight-sharing optimizations. We fully implement DiffoRA and design comprehensive experiments to evaluate its performance. The experimental results demonstrate that DiffoRA delivers state-of-the-art results across multiple benchmarks. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 02:41:34 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 07:14:31 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jiang",
"Tangyu",
""
],
[
"Wang",
"Haodi",
""
],
[
"Yuan",
"Chun",
""
]
] |
2502.09356 | Gabriel Tseng | Gabriel Tseng, Anthony Fuller, Marlena Reil, Henry Herzog, Patrick Beukema, Favyen Bastani, James R. Green, Evan Shelhamer, Hannah Kerner, David Rolnick | Galileo: Learning Global & Local Features of Many Remote Sensing Modalities | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a highly multimodal transformer to represent many remote sensing modalities - multispectral optical, synthetic aperture radar, elevation, weather, pseudo-labels, and more - across space and time. These inputs are useful for diverse remote sensing tasks, such as crop mapping and flood detection. However, learning shared representations of remote sensing data is challenging, given the diversity of relevant data modalities, and because objects of interest vary massively in scale, from small boats (1-2 pixels and fast) to glaciers (thousands of pixels and slow). We present a novel self-supervised learning algorithm that extracts multi-scale features across a flexible set of input modalities through masked modeling. Our dual global and local contrastive losses differ in their targets (deep representations vs. shallow input projections) and masking strategies (structured vs. not). Our Galileo is a single generalist model that outperforms SoTA specialist models for satellite images and pixel time series across eleven benchmarks and multiple tasks. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 14:21:03 GMT"
},
{
"version": "v2",
"created": "Wed, 28 May 2025 09:46:10 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 14:07:47 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Tseng",
"Gabriel",
""
],
[
"Fuller",
"Anthony",
""
],
[
"Reil",
"Marlena",
""
],
[
"Herzog",
"Henry",
""
],
[
"Beukema",
"Patrick",
""
],
[
"Bastani",
"Favyen",
""
],
[
"Green",
"James R.",
""
],
[
"Shelhamer",
"Evan",
""
],
[
"Kerner",
"Hannah",
""
],
[
"Rolnick",
"David",
""
]
] |
2502.09543 | Dmitry Shishkin | Noam Soker, Dmitry Shishkin | The main jet axis of the W49B supernova remnant | Accepted for publication in PASA | Publ. Astron. Soc. Aust. 42 (2025) e048 | 10.1017/pasa.2025.39 | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | We identify an axis connecting two opposite `ears' in the supernova remnant W49B and morphological signatures of three arcs around this axis that we claim are sections of full circum-jet rings. Based on recent identifications of morphological signatures of jets in core-collapse supernovae (CCSNe), including ejecta-rich axes, we reexamine images of W49B and identify a heavy element-rich protrusion (ear) as a jet-inflated structure. We identify the opposite ear and a clump at its tip as the signature of the opposite jets. The line connecting the two clumps at the tips of the two opposite ears forms the main jet axis of W49B. We compare the three arcs around the main jet axis in W49B to the circum-jet rings of the jets in the Cygnus A galaxy and deduce that these arcs are sections of full circum-jet rings in W49B. In W49B, the jets are long gone, as in some planetary nebulae with circum-jet rings. Identifying the main jet axis is incompatible with a type Ia supernova. It leaves two possibilities: that jets exploded W49B as a CCSN, i.e. the jittering jets explosion mechanism where the pair of jets we identify is one of many that exploded the star, or that the explosion was a common envelope jet supernova with a thermonuclear outburst, i.e., both the pair of jets and thermonuclear outburst exploded the core of a red supergiant star as a pre-existing neutron star tidally destroyed it. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 17:57:42 GMT"
},
{
"version": "v2",
"created": "Fri, 14 Feb 2025 13:59:26 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 18:18:44 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Soker",
"Noam",
""
],
[
"Shishkin",
"Dmitry",
""
]
] |
2502.09564 | Massimiliano Ciranni M.Sc. | Massimiliano Ciranni, Vito Paolo Pastore, Roberto Di Via, Enzo Tartaglione, Francesca Odone, Vittorio Murino | Diffusing DeBias: Synthetic Bias Amplification for Model Debiasing | 18 Pages, 9 Figures | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Deep learning model effectiveness in classification tasks is often challenged by the quality and quantity of training data whenever they are affected by strong spurious correlations between specific attributes and target labels. This results in a form of bias affecting training data, which typically leads to unrecoverable weak generalization in prediction. This paper aims at facing this problem by leveraging bias amplification with generated synthetic data: we introduce Diffusing DeBias (DDB), a novel approach acting as a plug-in for common methods of unsupervised model debiasing exploiting the inherent bias-learning tendency of diffusion models in data generation. Specifically, our approach adopts conditional diffusion models to generate synthetic bias-aligned images, which replace the original training set for learning an effective bias amplifier model that we subsequently incorporate into an end-to-end and a two-step unsupervised debiasing approach. By tackling the fundamental issue of bias-conflicting training samples memorization in learning auxiliary models, typical of this type of techniques, our proposed method beats current state-of-the-art in multiple benchmark datasets, demonstrating its potential as a versatile and effective tool for tackling bias in deep learning models. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 18:17:03 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Feb 2025 22:42:41 GMT"
},
{
"version": "v3",
"created": "Sun, 09 Mar 2025 18:41:50 GMT"
},
{
"version": "v4",
"created": "Wed, 04 Jun 2025 13:34:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ciranni",
"Massimiliano",
""
],
[
"Pastore",
"Vito Paolo",
""
],
[
"Di Via",
"Roberto",
""
],
[
"Tartaglione",
"Enzo",
""
],
[
"Odone",
"Francesca",
""
],
[
"Murino",
"Vittorio",
""
]
] |
2502.09591 | Chuanhui Liu | Chuanhui Liu and Xiao Wang | Censor Dependent Variational Inference | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | This paper provides a comprehensive analysis of variational inference in latent variable models for survival analysis, emphasizing the distinctive challenges associated with applying variational methods to survival data. We identify a critical weakness in the existing methodology, demonstrating how a poorly designed variational distribution may hinder the objective of survival analysis tasks - modeling time-to-event distributions. We prove that the optimal variational distribution, which perfectly bounds the log-likelihood, may depend on the censoring mechanism. To address this issue, we propose censor-dependent variational inference (CDVI), tailored for latent variable models in survival analysis. More practically, we introduce CD-CVAE, a V-structure Variational Autoencoder (VAE) designed for the scalable implementation of CDVI. Further discussion extends some existing theories and training techniques to survival analysis. Extensive experiments validate our analysis and demonstrate significant improvements in the estimation of individual survival distributions. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 18:48:04 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 22:24:30 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Liu",
"Chuanhui",
""
],
[
"Wang",
"Xiao",
""
]
] |
2502.09775 | Yuhui Zhang | Yuhui Zhang, Yuchang Su, Chenyu Wang, Tianhong Li, Zoe Wefers, Jeffrey Nirschl, James Burgess, Daisy Ding, Alejandro Lozano, Emma Lundberg, Serena Yeung-Levy | CellFlux: Simulating Cellular Morphology Changes via Flow Matching | Published at ICML 2025 | null | null | null | q-bio.QM cs.CV cs.LG q-bio.BM q-bio.CB | http://creativecommons.org/licenses/by/4.0/ | Building a virtual cell capable of accurately simulating cellular behaviors in silico has long been a dream in computational biology. We introduce CellFlux, an image-generative model that simulates cellular morphology changes induced by chemical and genetic perturbations using flow matching. Unlike prior methods, CellFlux models distribution-wise transformations from unperturbed to perturbed cell states, effectively distinguishing actual perturbation effects from experimental artifacts such as batch effects -- a major challenge in biological data. Evaluated on chemical (BBBC021), genetic (RxRx1), and combined perturbation (JUMP) datasets, CellFlux generates biologically meaningful cell images that faithfully capture perturbation-specific morphological changes, achieving a 35% improvement in FID scores and a 12% increase in mode-of-action prediction accuracy over existing methods. Additionally, CellFlux enables continuous interpolation between cellular states, providing a potential tool for studying perturbation dynamics. These capabilities mark a significant step toward realizing virtual cell modeling for biomedical research. Project page: https://yuhui-zh15.github.io/CellFlux/. | [
{
"version": "v1",
"created": "Thu, 13 Feb 2025 21:10:00 GMT"
},
{
"version": "v2",
"created": "Thu, 29 May 2025 00:19:28 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 20:03:58 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhang",
"Yuhui",
""
],
[
"Su",
"Yuchang",
""
],
[
"Wang",
"Chenyu",
""
],
[
"Li",
"Tianhong",
""
],
[
"Wefers",
"Zoe",
""
],
[
"Nirschl",
"Jeffrey",
""
],
[
"Burgess",
"James",
""
],
[
"Ding",
"Daisy",
""
],
[
"Lozano",
"Alejandro",
""
],
[
"Lundberg",
"Emma",
""
],
[
"Yeung-Levy",
"Serena",
""
]
] |
2502.10656 | Jian-Sheng Wang | Jian-Sheng Wang | Beyond the Drude model: surface and non-local effects in near-field radiative heat transfer and the Casimir puzzle | 12 pages 7 figures | Phys. Rev. B 111, 245404 (2025) | 10.1103/PhysRevB.111.245404 | null | cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the charge and current response functions $P$ and $\Pi$ in a semi-infinite metal block using the electron surface Green's functions. The surface electrons behave similarly to a two-dimensional Fermi gas but are strongly damped due to coupling to the bulk. This substantially reduces the region of validity of the Drude model for $P$, which requires the frequency $\omega \gg \max( v_F q, 1/\tau)$, here $v_F$ is the Fermi velocity, $q$ is the wavevector and $\tau$ is an effective relaxation time. As a consequence, for typical metal in near-field heat transfer, the Coulomb interaction goes as $1/d^4$ with the distance of the vacuum gap instead of the well-known $1/d^2$ of Drude model result. The current response $\Pi$ is shown to be highly anisotropic. The Drude model describes well the transverse directions parallel to the surface but is very different in the normal direction up to about 100 lattice sites away from the surface. These ideas and the residue diamagnetic effect of a nonzero $\Pi$ on the surface at zero frequency still cannot resolve the Casimir puzzle. | [
{
"version": "v1",
"created": "Sat, 15 Feb 2025 03:42:03 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wang",
"Jian-Sheng",
""
]
] |
2502.11673 | Adrian M\"uller | Adrian M\"uller and Jon Schneider and Stratis Skoulakis and Luca Viano and Volkan Cevher | Best of Both Worlds: Regret Minimization versus Minimax Play | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | In this paper, we investigate the existence of online learning algorithms with bandit feedback that simultaneously guarantee $O(1)$ regret compared to a given comparator strategy, and $\tilde{O}(\sqrt{T})$ regret compared to any fixed strategy, where $T$ is the number of rounds. We provide the first affirmative answer to this question whenever the comparator strategy supports every action. In the context of zero-sum games with min-max value zero, both in normal- and extensive form, we show that our results allow us to guarantee to risk at most $O(1)$ loss while being able to gain $\Omega(T)$ from exploitable opponents, thereby combining the benefits of both no-regret algorithms and minimax play. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 11:04:01 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 09:29:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Müller",
"Adrian",
""
],
[
"Schneider",
"Jon",
""
],
[
"Skoulakis",
"Stratis",
""
],
[
"Viano",
"Luca",
""
],
[
"Cevher",
"Volkan",
""
]
] |
2502.11753 | Michiel Van Der Meer | Michiel van der Meer, Pavel Korshunov, S\'ebastien Marcel, Lonneke van der Plas | HintsOfTruth: A Multimodal Checkworthiness Detection Dataset with Real and Synthetic Claims | Accepted at ACL2025 (main track) | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Misinformation can be countered with fact-checking, but the process is costly and slow. Identifying checkworthy claims is the first step, where automation can help scale fact-checkers' efforts. However, detection methods struggle with content that is (1) multimodal, (2) from diverse domains, and (3) synthetic. We introduce HintsOfTruth, a public dataset for multimodal checkworthiness detection with 27K real-world and synthetic image/claim pairs. The mix of real and synthetic data makes this dataset unique and ideal for benchmarking detection methods. We compare fine-tuned and prompted Large Language Models (LLMs). We find that well-configured lightweight text-based encoders perform comparably to multimodal models but the former only focus on identifying non-claim-like content. Multimodal LLMs can be more accurate but come at a significant computational cost, making them impractical for large-scale applications. When faced with synthetic data, multimodal models perform more robustly. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 12:49:55 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 11:13:28 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"van der Meer",
"Michiel",
""
],
[
"Korshunov",
"Pavel",
""
],
[
"Marcel",
"Sébastien",
""
],
[
"van der Plas",
"Lonneke",
""
]
] |
2502.12010 | Thiago Fassarella T.Fassarella | Thiago Fassarella and Nivaldo Medeiros | Hyperplane arrangements and the Gauss map of a pencil | null | null | null | null | math.AG math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show that the coefficients of the characteristic polynomial of a central hyperplane arrangement $\mathcal A$, coincide with the multidegrees of the Gauss map of a pencil of hypersurfaces naturally associated to $\mathcal A$. As a consequence, we obtain a proof of the Heron-Rota-Welsh conjecture for matroids representable over a field of characteristic zero. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 16:44:11 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 17:43:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Fassarella",
"Thiago",
""
],
[
"Medeiros",
"Nivaldo",
""
]
] |
2502.12031 | Aurian Quelennec | Aurian Quelennec, Pierre Chouteau, Geoffroy Peeters, Slim Essid | Masked Latent Prediction and Classification for Self-Supervised Audio Representation Learning | Copyright 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works | ICASSP 2025 | 10.1109/ICASSP49660.2025.10887666 | null | cs.SD cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, self-supervised learning methods based on masked latent prediction have proven to encode input data into powerful representations. However, during training, the learned latent space can be further transformed to extract higher-level information that could be more suited for downstream classification tasks. Therefore, we propose a new method: MAsked latenT Prediction And Classification (MATPAC), which is trained with two pretext tasks solved jointly. As in previous work, the first pretext task is a masked latent prediction task, ensuring a robust input representation in the latent space. The second one is unsupervised classification, which utilises the latent representations of the first pretext task to match probability distributions between a teacher and a student. We validate the MATPAC method by comparing it to other state-of-the-art proposals and conducting ablations studies. MATPAC reaches state-of-the-art self-supervised learning results on reference audio classification datasets such as OpenMIC, GTZAN, ESC-50 and US8K and outperforms comparable supervised methods results for musical auto-tagging on Magna-tag-a-tune. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 17:02:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Quelennec",
"Aurian",
""
],
[
"Chouteau",
"Pierre",
""
],
[
"Peeters",
"Geoffroy",
""
],
[
"Essid",
"Slim",
""
]
] |
2502.12089 | Alessandro Favero | Alessandro Favero, Antonio Sclocchi, Francesco Cagnetta, Pascal Frossard and Matthieu Wyart | How Compositional Generalization and Creativity Improve as Diffusion Models are Trained | null | Proceedings of the 42nd International Conference on Machine Learning (ICML), PMLR 267, 2025 | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Natural data is often organized as a hierarchical composition of features. How many samples do generative models need in order to learn the composition rules, so as to produce a combinatorially large number of novel data? What signal in the data is exploited to learn those rules? We investigate these questions in the context of diffusion models both theoretically and empirically. Theoretically, we consider a simple probabilistic context-free grammar - a tree-like graphical model used to represent the hierarchical and compositional structure of data such as language and images. We demonstrate that diffusion models learn the grammar's composition rules with the sample complexity required for clustering features with statistically similar context, a process similar to the word2vec algorithm. However, this clustering emerges hierarchically: higher-level features associated with longer contexts require more data to be identified. This mechanism leads to a sample complexity that scales polynomially with the said context size. As a result, diffusion models trained on an intermediate dataset size generate data coherent up to a certain scale, but lacking global coherence. We test these predictions across different domains and find remarkable agreement: both generated texts and images achieve progressively larger coherence lengths as the training time or dataset size grows. We discuss connections between the hierarchical clustering mechanism we introduce here and the renormalization group in physics. | [
{
"version": "v1",
"created": "Mon, 17 Feb 2025 18:06:33 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Mar 2025 20:57:35 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 11:03:45 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Favero",
"Alessandro",
""
],
[
"Sclocchi",
"Antonio",
""
],
[
"Cagnetta",
"Francesco",
""
],
[
"Frossard",
"Pascal",
""
],
[
"Wyart",
"Matthieu",
""
]
] |
2502.12708 | Jean-Philip Piquemal | Nastasia Mauger, Thomas Pl\'e, Louis Lagard\`ere, Simon Huppert, Jean-Philip Piquemal | The Q-AMOEBA (CF) Polarizable Potential | null | J. Phys. Chem. Lett., 2025, 16, 5723-5731 | 10.1021/acs.jpclett.5c00797 | null | physics.chem-ph | http://creativecommons.org/licenses/by/4.0/ | We present Q-AMOEBA (CF), an enhanced version of the Q-AMOEBA polarizable model that integrates a geometry-dependent charge flux (CF) term while designed for an explicit treatment of nuclear quantum effects (NQE). The inclusion of CF effects allows matching experimental data for the molecular structure of water in both gas and liquid phases, addressing limitations faced by most force fields. We show that Q-AMOEBA (CF) provides highly accurate results for a wide range of thermodynamical properties of liquid water. Using the computational efficiency of the adaptive Quantum Thermal Bath method, which accounts for NQE at a cost comparable to classical molecular dynamics, we evaluate the robustness and transferability of Q-AMOEBA (CF) by calculating hydration free energies of various ions and organic molecules. Finally, we apply this methodology to the alanine dipeptide and compute the corresponding dihedral angle potential of mean force and hydration free energy. Unexpectedly, the latter quantity displays significant NQE. These results pave the way to a finer understanding of their role in biochemical systems. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 10:20:57 GMT"
},
{
"version": "v2",
"created": "Sat, 15 Mar 2025 18:47:30 GMT"
},
{
"version": "v3",
"created": "Sun, 11 May 2025 17:09:36 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Mauger",
"Nastasia",
""
],
[
"Plé",
"Thomas",
""
],
[
"Lagardère",
"Louis",
""
],
[
"Huppert",
"Simon",
""
],
[
"Piquemal",
"Jean-Philip",
""
]
] |
2502.13135 | Taedong Yun | Taedong Yun, Eric Yang, Mustafa Safdari, Jong Ha Lee, Vaishnavi Vinod Kumar, S. Sara Mahdavi, Jonathan Amar, Derek Peyton, Reut Aharony, Andreas Michaelides, Logan Schneider, Isaac Galatzer-Levy, Yugang Jia, John Canny, Arthur Gretton, Maja Matari\'c | Sleepless Nights, Sugary Days: Creating Synthetic Users with Health Conditions for Realistic Coaching Agent Interactions | Accepted to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025) | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | We present an end-to-end framework for generating synthetic users for evaluating interactive agents designed to encourage positive behavior changes, such as in health and lifestyle coaching. The synthetic users are grounded in health and lifestyle conditions, specifically sleep and diabetes management in this study, to ensure realistic interactions with the health coaching agent. Synthetic users are created in two stages: first, structured data are generated grounded in real-world health and lifestyle factors in addition to basic demographics and behavioral attributes; second, full profiles of the synthetic users are developed conditioned on the structured data. Interactions between synthetic users and the coaching agent are simulated using generative agent-based models such as Concordia, or directly by prompting a language model. Using two independently-developed agents for sleep and diabetes coaching as case studies, the validity of this framework is demonstrated by analyzing the coaching agent's understanding of the synthetic users' needs and challenges. Finally, through multiple blinded evaluations of user-coach interactions by human experts, we demonstrate that our synthetic users with health and behavioral attributes more accurately portray real human users with the same attributes, compared to generic synthetic users not grounded in such attributes. The proposed framework lays the foundation for efficient development of conversational agents through extensive, realistic, and grounded simulated interactions. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 18:56:44 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 14:50:52 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yun",
"Taedong",
""
],
[
"Yang",
"Eric",
""
],
[
"Safdari",
"Mustafa",
""
],
[
"Lee",
"Jong Ha",
""
],
[
"Kumar",
"Vaishnavi Vinod",
""
],
[
"Mahdavi",
"S. Sara",
""
],
[
"Amar",
"Jonathan",
""
],
[
"Peyton",
"Derek",
""
],
[
"Aharony",
"Reut",
""
],
[
"Michaelides",
"Andreas",
""
],
[
"Schneider",
"Logan",
""
],
[
"Galatzer-Levy",
"Isaac",
""
],
[
"Jia",
"Yugang",
""
],
[
"Canny",
"John",
""
],
[
"Gretton",
"Arthur",
""
],
[
"Matarić",
"Maja",
""
]
] |
2502.13302 | Philip Richerme | Philip Richerme | Multi-Mode Global Driving of Trapped Ions for Quantum Circuit Synthesis | 23 pages, 10 figures | null | null | null | quant-ph physics.atom-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the use of global drives with multiple frequency components to improve the efficiency of trapped ion quantum simulations and computations. We show that such `multi-mode' global drives, when combined with a linear number of single-qubit rotations, generate universal Ising-type interactions with shorter overall runtimes than corresponding two-qubit gate implementations. Further, we show how this framework may be extended to efficiently generate $n-$body interactions between any subset $n$ of the ion qubits. Finally, we apply these techniques to encode the Quantum Fourier Transform using quadratically-fewer entangling operations, with quadratically smaller runtime, compared with traditional approaches. | [
{
"version": "v1",
"created": "Tue, 18 Feb 2025 21:47:51 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 17:49:32 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Richerme",
"Philip",
""
]
] |
2502.13656 | Liyang He | Liyang He, Chenglong Liu, Rui Li, Zhenya Huang, Shulan Ruan, Jun Zhou, Enhong Chen | Refining Sentence Embedding Model through Ranking Sentences Generation with Large Language Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Sentence embedding is essential for many NLP tasks, with contrastive learning methods achieving strong performance using annotated datasets like NLI. Yet, the reliance on manual labels limits scalability. Recent studies leverage large language models (LLMs) to generate sentence pairs, reducing annotation dependency. However, they overlook ranking information crucial for fine-grained semantic distinctions. To tackle this challenge, we propose a method for controlling the generation direction of LLMs in the latent space. Unlike unconstrained generation, the controlled approach ensures meaningful semantic divergence. Then, we refine exist sentence embedding model by integrating ranking information and semantic information. Experiments on multiple benchmarks demonstrate that our method achieves new SOTA performance with a modest cost in ranking sentence synthesis. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 12:07:53 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 16:39:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"He",
"Liyang",
""
],
[
"Liu",
"Chenglong",
""
],
[
"Li",
"Rui",
""
],
[
"Huang",
"Zhenya",
""
],
[
"Ruan",
"Shulan",
""
],
[
"Zhou",
"Jun",
""
],
[
"Chen",
"Enhong",
""
]
] |
2502.13683 | Madhu Mishra | Soumya Adhikari, Aravind Aikot, Bindusar Sahoo, and Madhu Mishra | Variant dilaton Weyl Multiplet for N=3 conformal supergravity in four dimensions | 16 pages, added new references | null | 10.1103/PhysRevD.111.106024 | null | hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We construct a new dilaton Weyl multiplet for $\mathcal{N}=3$ conformal supergravity in four dimensions. The R-symmetry realized on this dilaton Weyl multiplet is $SU(2) \times U(1) \times U(1)$. The construction follows a two-step procedure. Firstly, two on-shell vector multiplets are coupled to the standard Weyl multiplet. Secondly, using the field equations of the vector multiplets, some of the auxiliary fields of the standard Weyl multiplet are solved in terms of the fields belonging to the vector multiplets and some dual gauge fields. The remaining fields of the standard Weyl multiplet combine with the vector multiplet fields and the dual gauge fields to constitute the new dilaton Weyl multiplet. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 12:51:58 GMT"
},
{
"version": "v2",
"created": "Tue, 04 Mar 2025 07:17:46 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Adhikari",
"Soumya",
""
],
[
"Aikot",
"Aravind",
""
],
[
"Sahoo",
"Bindusar",
""
],
[
"Mishra",
"Madhu",
""
]
] |
2502.13946 | Chak Tou Leong | Chak Tou Leong, Qingyu Yin, Jian Wang, Wenjie Li | Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region | ACL 2025 Main | null | null | null | cs.CL cs.AI cs.CR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The safety alignment of large language models (LLMs) remains vulnerable, as their initial behavior can be easily jailbroken by even relatively simple attacks. Since infilling a fixed template between the input instruction and initial model output is a common practice for existing LLMs, we hypothesize that this template is a key factor behind their vulnerabilities: LLMs' safety-related decision-making overly relies on the aggregated information from the template region, which largely influences these models' safety behavior. We refer to this issue as template-anchored safety alignment. In this paper, we conduct extensive experiments and verify that template-anchored safety alignment is widespread across various aligned LLMs. Our mechanistic analyses demonstrate how it leads to models' susceptibility when encountering inference-time jailbreak attacks. Furthermore, we show that detaching safety mechanisms from the template region is promising in mitigating vulnerabilities to jailbreak attacks. We encourage future research to develop more robust safety alignment techniques that reduce reliance on the template region. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 18:42:45 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:20:11 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Leong",
"Chak Tou",
""
],
[
"Yin",
"Qingyu",
""
],
[
"Wang",
"Jian",
""
],
[
"Li",
"Wenjie",
""
]
] |
2502.13967 | Roman Bachmann | Roman Bachmann, Jesse Allardice, David Mizrahi, Enrico Fini, O\u{g}uzhan Fatih Kar, Elmira Amirloo, Alaaeldin El-Nouby, Amir Zamir, Afshin Dehghan | FlexTok: Resampling Images into 1D Token Sequences of Flexible Length | ICML 2025. Project page at https://flextok.epfl.ch/ | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Image tokenization has enabled major advances in autoregressive image generation by providing compressed, discrete representations that are more efficient to process than raw pixels. While traditional approaches use 2D grid tokenization, recent methods like TiTok have shown that 1D tokenization can achieve high generation quality by eliminating grid redundancies. However, these methods typically use a fixed number of tokens and thus cannot adapt to an image's inherent complexity. We introduce FlexTok, a tokenizer that projects 2D images into variable-length, ordered 1D token sequences. For example, a 256x256 image can be resampled into anywhere from 1 to 256 discrete tokens, hierarchically and semantically compressing its information. By training a rectified flow model as the decoder and using nested dropout, FlexTok produces plausible reconstructions regardless of the chosen token sequence length. We evaluate our approach in an autoregressive generation setting using a simple GPT-style Transformer. On ImageNet, this approach achieves an FID<2 across 8 to 128 tokens, outperforming TiTok and matching state-of-the-art methods with far fewer tokens. We further extend the model to support to text-conditioned image generation and examine how FlexTok relates to traditional 2D tokenization. A key finding is that FlexTok enables next-token prediction to describe images in a coarse-to-fine "visual vocabulary", and that the number of tokens to generate depends on the complexity of the generation task. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 18:59:44 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Jun 2025 12:56:23 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bachmann",
"Roman",
""
],
[
"Allardice",
"Jesse",
""
],
[
"Mizrahi",
"David",
""
],
[
"Fini",
"Enrico",
""
],
[
"Kar",
"Oğuzhan Fatih",
""
],
[
"Amirloo",
"Elmira",
""
],
[
"El-Nouby",
"Alaaeldin",
""
],
[
"Zamir",
"Amir",
""
],
[
"Dehghan",
"Afshin",
""
]
] |
2502.14019 | Myra Cheng | Myra Cheng, Su Lin Blodgett, Alicia DeVrio, Lisa Egede, Alexandra Olteanu | Dehumanizing Machines: Mitigating Anthropomorphic Behaviors in Text Generation Systems | ACL 2025 | null | null | null | cs.CL cs.AI cs.HC | http://creativecommons.org/licenses/by/4.0/ | As text generation systems' outputs are increasingly anthropomorphic -- perceived as human-like -- scholars have also increasingly raised concerns about how such outputs can lead to harmful outcomes, such as users over-relying or developing emotional dependence on these systems. How to intervene on such system outputs to mitigate anthropomorphic behaviors and their attendant harmful outcomes, however, remains understudied. With this work, we aim to provide empirical and theoretical grounding for developing such interventions. To do so, we compile an inventory of interventions grounded both in prior literature and a crowdsourcing study where participants edited system outputs to make them less human-like. Drawing on this inventory, we also develop a conceptual framework to help characterize the landscape of possible interventions, articulate distinctions between different types of interventions, and provide a theoretical basis for evaluating the effectiveness of different interventions. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 18:06:37 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Jun 2025 18:14:14 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cheng",
"Myra",
""
],
[
"Blodgett",
"Su Lin",
""
],
[
"DeVrio",
"Alicia",
""
],
[
"Egede",
"Lisa",
""
],
[
"Olteanu",
"Alexandra",
""
]
] |
2502.14157 | Xing Fan | Xing Fan, Atsushi Noguchi, and Kento Taniguchi | Quantum Logic Spectroscopy of an Electron and Positron for Precise Tests of the Standard Model | 8 pages, 4 figures | Phys. Rev. A 111, 042806 (2025) | 10.1103/PhysRevA.111.042806 | null | quant-ph hep-ex physics.atom-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a scheme for quantum logic spectroscopy of an electron or positron in a Penning trap. An electron or positron in a spectroscopy trap is coupled to a remote logic electron or positron via a wire to achieve motional entanglement. By separating the two traps, one can significantly improve magnetic field homogeneity, microwave characteristics, and detection sensitivity. The proposed scheme will improve the measurement precision of the electron's and positron's magnetic moments and charge-to-mass ratios, enabling precise tests of the Standard Model of particle physics. | [
{
"version": "v1",
"created": "Wed, 19 Feb 2025 23:53:38 GMT"
},
{
"version": "v2",
"created": "Sat, 22 Mar 2025 11:56:27 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jun 2025 00:07:54 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Fan",
"Xing",
""
],
[
"Noguchi",
"Atsushi",
""
],
[
"Taniguchi",
"Kento",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.