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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2404.16085 | Sihao Cheng | Sihao Cheng, Gabriela A. Marques, Daniela Grand\'on, Leander Thiele, Masato Shirasaki, Brice M\'enard, Jia Liu | Cosmological constraints from weak lensing scattering transform using HSC Y1 data | 8 pages, 6 figures | Journal of Cosmology and Astroparticle Physics, Volume 2025, Issue 01, No. 006 | 10.1088/1475-7516/2025/01/006 | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | As weak lensing surveys go deeper, there is an increasing need for reliable characterization of non-Gaussian structures at small angular scales. Here we present the first cosmological constraints with weak lensing scattering transform, a statistical estimator that combines efficiency, robustness, and interpretability. With the Hyper Suprime-Cam survey (HSC) year 1 data, we obtain $\Omega_\text{m}=0.29_{-0.03}^{+0.04}$, $S_8\equiv \sigma_8(\Omega_\text{m}/0.3)^{0.5}=0.83\pm0.02$, and intrinsic alignment strength $A_\text{IA}=1.0\pm0.4$ through simulation-based forward modeling. Our constraints are consistent with those derived from Planck. The error bar of $\Omega_\text{m}$ is 2 times tighter than that obtained from the power spectrum when the same scale range is used. This constraining power is on par with that of convolutional neural networks, suggesting that further investment in spatial information extraction may not yield substantial benefits.
We also point out an internal tension of $S_8$ estimates linked to a redshift bin around z ~ 1 in the HSC data. We found that discarding that bin leads to a consistent decrease of $S_8$ from 0.83 to 0.79, for all statistical estimators. We argue that photometric redshift estimation is now the main limitation in the estimation of $S_8$ using HSC. This limitation is likely to affect other ground-based weak lensing surveys reaching redshifts greater than one. Alternative redshift estimation techniques, like clustering redshifts, may help alleviate this limitation. | [
{
"version": "v1",
"created": "Wed, 24 Apr 2024 18:00:00 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Cheng",
"Sihao",
""
],
[
"Marques",
"Gabriela A.",
""
],
[
"Grandón",
"Daniela",
""
],
[
"Thiele",
"Leander",
""
],
[
"Shirasaki",
"Masato",
""
],
[
"Ménard",
"Brice",
""
],
[
"Liu",
"Jia",
""
]
] |
2405.02437 | Abdulrahman Diaa | Abdulrahman Diaa, Thomas Humphries, Florian Kerschbaum | FastLloyd: Federated, Accurate, Secure, and Tunable $k$-Means Clustering with Differential Privacy | In 34th Usenix Security Symposium | null | null | null | cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problem of privacy-preserving $k$-means clustering in the horizontally federated setting. Existing federated approaches using secure computation suffer from substantial overheads and do not offer output privacy. At the same time, differentially private (DP) $k$-means algorithms either assume a trusted central curator or significantly degrade utility by adding noise in the local DP model. Naively combining the secure and central DP solutions results in a protocol with impractical overhead. Instead, our work provides enhancements to both the DP and secure computation components, resulting in a design that is faster, more private, and more accurate than previous work. By utilizing the computational DP model, we design a lightweight, secure aggregation-based approach that achieves five orders of magnitude speed-up over state-of-the-art related work. Furthermore, we not only maintain the utility of the state-of-the-art in the central model of DP, but we improve the utility further by designing a new DP clustering mechanism. | [
{
"version": "v1",
"created": "Fri, 03 May 2024 19:04:37 GMT"
},
{
"version": "v2",
"created": "Tue, 01 Apr 2025 15:45:51 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:45:50 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Diaa",
"Abdulrahman",
""
],
[
"Humphries",
"Thomas",
""
],
[
"Kerschbaum",
"Florian",
""
]
] |
2405.04742 | Martin Kuffer | Martin Kuffer, Anal\'ia Zwick, Gonzalo A. \'Alvarez | Sensing Out-of-Equilibrium and Quantum Non-Gaussian environments via induced Time-Reversal Symmetry Breaking on the quantum-probe dynamics | null | PRX Quantum 6, 020320 (2025) | 10.1103/PRXQuantum.6.020320 | null | quant-ph cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Advancing quantum sensing tools for investigating systems at atomic and nanoscales is crucial for the progress of quantum technologies. While numerous protocols employ quantum probes to extract information from stationary or weakly coupled environments, the challenges intensify at atomic- and nano-scales where the environment is inherently out-of-equilibrium or strongly coupled with the sensor. We here prove that the time-reversal symmetry in the quantum-sensor control dynamics is broken, when partial information is probed from an environment that is out-of-equilibrium with non stationary fluctuations or is described by quantum non-Gaussian, strongly coupled environmental correlations. We exploit this phenomenon as a quantum sensing paradigm with proof-of principle experimental quantum simulations using solid-state nuclear magnetic resonance (NMR). This introduces a signal contrast on a qubit-probe that quantifies how far the sensed environment is from equilibrium or its quantum non-Gaussian nature. Protocols are also presented to discern and filter a variety of environmental properties including stationary, non-stationary and non-Gaussian quantum noise fluctuations as a step toward sensing the ubiquitous environments of a quantum-sensor at atomic and nanoscales. | [
{
"version": "v1",
"created": "Wed, 08 May 2024 01:13:07 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kuffer",
"Martin",
""
],
[
"Zwick",
"Analía",
""
],
[
"Álvarez",
"Gonzalo A.",
""
]
] |
2405.05132 | Yi-Jun Chang | Yi-Jun Chang and Varsha Dani and Thomas P. Hayes | Low-Distortion Clustering in Bounded Growth Graphs | null | null | null | null | cs.DC cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The well-known clustering algorithm of Miller, Peng, and Xu (SPAA 2013) is useful for many applications, including low-diameter decomposition and low-energy distributed algorithms. One nice property of their clustering, shown in previous work by Chang, Dani, Hayes, and Pettie (PODC 2020), is that distances in the cluster graph are rescaled versions of distances in the original graph, up to an $O(\log n)$ distortion factor and rounding issues. Minimizing this distortion factor is important for efficiency in computing the clustering, as well as in further applications, once the clustering has been constructed.
We prove that there exist graphs for which an $\Omega((\log n)^{1/3})$ distortion factor is necessary for any clustering. We also consider a class of nice graphs which we call uniformly bounded independence graphs. These include, for example, paths, lattice graphs, and "dense" unit disk graphs. For these graphs, we prove that clusterings of constant distortion always exist, and moreover, we give an efficient distributed algorithm to construct them. Our clustering algorithm is based on Voronoi cells centered at the vertices of a maximal independent set in a suitable power graph.
Applications of our new clustering include low-energy simulation of distributed algorithms in the LOCAL, CONGEST, and RADIO-CONGEST models, as well as efficient approximate solutions to distributed combinatorial optimization problems. We complement these results with matching or nearly matching lower bounds. | [
{
"version": "v1",
"created": "Wed, 08 May 2024 15:31:30 GMT"
},
{
"version": "v2",
"created": "Fri, 06 Sep 2024 04:57:11 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 04:07:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chang",
"Yi-Jun",
""
],
[
"Dani",
"Varsha",
""
],
[
"Hayes",
"Thomas P.",
""
]
] |
2405.05233 | Zhongqi Liang | Zhongqi Liang, Jes\'us P\'erez-R\'ios | Classical Grand Angular Momentum in N-Body Problems | 14 pages, 3 figures | Few-Body Syst. 66, 25 (2025) | 10.1007/s00601-025-01996-8 | null | math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | The concept of grand angular momentum is widely used in the study of N-body problems quantum mechanically. Here, we applied it to a classical analysis of N-body problems. Utilizing the tree representation for Jacobi and hyperspherical coordinates, we found a decomposition of its magnitude into magnitudes of one-body angular momenta in three dimensions. We generalized some results from the two-body case and derived a general expression for the scattering angle in N-body problems. | [
{
"version": "v1",
"created": "Wed, 08 May 2024 17:27:16 GMT"
},
{
"version": "v2",
"created": "Thu, 03 Apr 2025 15:15:19 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Liang",
"Zhongqi",
""
],
[
"Pérez-Ríos",
"Jesús",
""
]
] |
2405.05434 | Giuseppe Vacca | L. Beir\~ao da Veiga, F. Dassi, G. Vacca | Pressure and convection robust Finite Elements for Magnetohydrodynamics | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose and analyze two convection quasi-robust and pressure robust finite element methods for a fully nonlinear time-dependent magnetohydrodynamics problem. Both methods employ the $H_{\rm div}$ conforming BDM element coupled with an appropriate pressure space guaranteeing the exact diagram for the fluid part, and the $H^1$ conforming Lagrange element for the approximation of the magnetic fluxes, and make use of suitable DG upwind terms and CIP stabilizations to handle the fluid and magnetic convective terms.
The main difference between the two approaches here proposed (labeled as three-field scheme and four field-scheme respectively) lies in the strategy adopted to enforce the divergence-free condition of the magnetic field.
The three-filed scheme implements a grad-div stabilization, whereas the four-field scheme introduces a suitable Lagrange multiplier and additional stabilization terms in the formulation.
The developed error estimates for the two schemes are uniform in both diffusion parameters and optimal with respect to the diffusive norm. Furthermore, in the convection dominated regime, being $k$ the degree of the method and $h$ the mesh size, we are able to prove $O(h^k)$ and $O(h^{k+1/2})$ pre-asymptotic error reduction rate for the three-field scheme and four-filed scheme respectively.
A set of numerical tests support our theoretical findings. | [
{
"version": "v1",
"created": "Wed, 08 May 2024 21:30:17 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 02:38:41 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"da Veiga",
"L. Beirão",
""
],
[
"Dassi",
"F.",
""
],
[
"Vacca",
"G.",
""
]
] |
2405.05946 | Luka Antonic | Luka Antonic, Yariv Kafri, Daniel Podolsky, Ari M. Turner | Motion from Measurement: The Role of Symmetry of Quantum Measurements | 20 pages, 11 figures | Phys. Rev. B 111, 224305 (2025) | 10.1103/PhysRevB.111.224305 | null | quant-ph cond-mat.other cond-mat.stat-mech | http://creativecommons.org/licenses/by/4.0/ | In quantum mechanics, measurements are dynamical processes and thus they should be capable of inducing currents. The symmetries of the Hamiltonian and measurement operator provide an organizing principle for understanding the conditions for such currents to emerge. The central role is played by the inversion and time-reversal symmetries. We classify the distinct behaviors that emerge from single and repeated measurements, with and without coupling to a dissipative bath. While the breaking of inversion symmetry alone is sufficient to generate currents through measurements, the breaking of time-reversal symmetry by the measurement operator leads to a dramatic increase in the magnitude of the currents. We consider the dependence on the measurement rate and find that the current is non-monotonic. Furthermore, nondegenerate measurements can lead to current loops within the steady state even in the Zeno limit. | [
{
"version": "v1",
"created": "Thu, 09 May 2024 17:35:49 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Antonic",
"Luka",
""
],
[
"Kafri",
"Yariv",
""
],
[
"Podolsky",
"Daniel",
""
],
[
"Turner",
"Ari M.",
""
]
] |
2405.07366 | Kevin Abela Mr | Abela Kevin, Chetcuti Emmanuel | Unbounded order convergence on infinitely distributive lattices | null | null | null | null | math.FA | http://creativecommons.org/licenses/by/4.0/ | We study uO convergence on infinitely distributive lattices, extending key properties known from Riesz spaces. We show that order continuity of uO convergence characterizes infinite distributivity. We examine O-adherence and uO adherence of sublattices and ideals, proving that the uO and O closures of a sublattice coincide and form a sublattice, and that the first uO adherence of an ideal is an O closed ideal. We also analyze the Dedekind MacNeille completion of a sublattice Y within that of a lattice L, identifying conditions (A) and (B) under which the completion of Y embeds regularly in that of L. In this case, we show that the first uO adherence of Y covers its O closure. | [
{
"version": "v1",
"created": "Sun, 12 May 2024 19:42:25 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Jun 2024 17:45:01 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 19:52:03 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kevin",
"Abela",
""
],
[
"Emmanuel",
"Chetcuti",
""
]
] |
2405.07763 | Yi Zhao | Ayush Basu, Vojtech Rodl, and Yi Zhao | The number of cliques in hypergraphs with forbidden subgraphs | null | Discrete Mathematics 348 (2025), Paper No. 114415, 7pp | 10.1016/j.disc.2025.114415 | null | math.CO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We study the maximum number of $r$-vertex cliques in $(r-1)$-uniform hypergraphs not containing complete $r$-partite hypergraphs $K_r^{(r-1)}(a_1, \dots, a_r)$. By using the hypergraph removal lemma, we show that this maximum is $o( n^{r - 1/(a_1 \cdots a_{r-1})} )$. This immediately implies the corresponding results of Mubayi and Mukherjee and of Balogh, Jiang, and Luo for graphs. We also provide a lower bound by using hypergraph Tur\'an numbers. | [
{
"version": "v1",
"created": "Mon, 13 May 2024 14:05:36 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:49:13 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Basu",
"Ayush",
""
],
[
"Rodl",
"Vojtech",
""
],
[
"Zhao",
"Yi",
""
]
] |
2405.11985 | Jingqun Tang | Jingqun Tang, Qi Liu, Yongjie Ye, Jinghui Lu, Shu Wei, Chunhui Lin, Wanqing Li, Mohamad Fitri Faiz Bin Mahmood, Hao Feng, Zhen Zhao, Yangfan He, Kuan Lu, Yanjie Wang, Yuliang Liu, Hao Liu, Xiang Bai, Can Huang | MTVQA: Benchmarking Multilingual Text-Centric Visual Question Answering | Accepted by ACL 2025 findings | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Text-Centric Visual Question Answering (TEC-VQA) in its proper format not only facilitates human-machine interaction in text-centric visual environments but also serves as a de facto gold proxy to evaluate AI models in the domain of text-centric scene understanding. Nonetheless, most existing TEC-VQA benchmarks have focused on high-resource languages like English and Chinese. Despite pioneering works to expand multilingual QA pairs in non-text-centric VQA datasets through translation engines, the translation-based protocol encounters a substantial "visual-textual misalignment" problem when applied to TEC-VQA. Specifically, it prioritizes the text in question-answer pairs while disregarding the visual text present in images. Moreover, it fails to address complexities related to nuanced meaning, contextual distortion, language bias, and question-type diversity. In this work, we tackle multilingual TEC-VQA by introducing MTVQA, the first benchmark featuring high-quality human expert annotations across 9 diverse languages, consisting of 6,778 question-answer pairs across 2,116 images. Further, by comprehensively evaluating numerous state-of-the-art Multimodal Large Language Models~(MLLMs), including Qwen2-VL, GPT-4o, GPT-4V, Claude3, and Gemini, on the MTVQA benchmark, it is evident that there is still a large room for performance improvement (Qwen2-VL scoring 30.9 versus 79.7 for human performance), underscoring the value of MTVQA. Additionally, we supply multilingual training data within the MTVQA dataset, demonstrating that straightforward fine-tuning with this data can substantially enhance multilingual TEC-VQA performance. We aspire that MTVQA will offer the research community fresh insights and stimulate further exploration in multilingual visual text comprehension. The project homepage is available at https://bytedance.github.io/MTVQA/. | [
{
"version": "v1",
"created": "Mon, 20 May 2024 12:35:01 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Jun 2024 09:32:56 GMT"
},
{
"version": "v3",
"created": "Tue, 19 Nov 2024 07:59:39 GMT"
},
{
"version": "v4",
"created": "Wed, 28 May 2025 03:52:29 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 03:33:02 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tang",
"Jingqun",
""
],
[
"Liu",
"Qi",
""
],
[
"Ye",
"Yongjie",
""
],
[
"Lu",
"Jinghui",
""
],
[
"Wei",
"Shu",
""
],
[
"Lin",
"Chunhui",
""
],
[
"Li",
"Wanqing",
""
],
[
"Mahmood",
"Mohamad Fitri Faiz Bin",
""
],
[
"Feng",
"Hao",
""
],
[
"Zhao",
"Zhen",
""
],
[
"He",
"Yangfan",
""
],
[
"Lu",
"Kuan",
""
],
[
"Wang",
"Yanjie",
""
],
[
"Liu",
"Yuliang",
""
],
[
"Liu",
"Hao",
""
],
[
"Bai",
"Xiang",
""
],
[
"Huang",
"Can",
""
]
] |
2405.14259 | Chan-Jan Hsu | Chan-Jan Hsu, Yi-Chang Chen, Feng-Ting Liao, Pei-Chen Ho, Yu-Hsiang Wang, Po-Chun Hsu, Da-shan Shiu | Let's Fuse Step by Step: A Generative Fusion Decoding Algorithm with LLMs for Robust and Instruction-Aware ASR and OCR | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | We propose "Generative Fusion Decoding" (GFD), a novel shallow fusion framework designed to integrate large language models (LLMs) into cross-modal text recognition systems for automatic speech recognition (ASR) and optical character recognition (OCR). We derive the necessary formulations to enable GFD to operate across mismatched token spaces of different models by calculating likelihood at the byte level, thereby enabling seamless fusion and synchronous progression during the decoding process. GFD is plug-and-play by design, making it readily compatible with various auto-regressive models without the need for any re-training. GFD proves effective for general ASR and OCR tasks through intermediate and frequent interactions with LLMs, surpassing cascaded methods in English and Mandarin benchmarks. In addition, GFD transfers in-context learning abilities of LLMs and allows for adaptive ASR in instruction-aware and long-context settings, yielding significant WER reductions of up to 17.7\%. | [
{
"version": "v1",
"created": "Thu, 23 May 2024 07:39:42 GMT"
},
{
"version": "v2",
"created": "Tue, 28 May 2024 14:45:30 GMT"
},
{
"version": "v3",
"created": "Sun, 02 Jun 2024 16:30:00 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 17:09:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Hsu",
"Chan-Jan",
""
],
[
"Chen",
"Yi-Chang",
""
],
[
"Liao",
"Feng-Ting",
""
],
[
"Ho",
"Pei-Chen",
""
],
[
"Wang",
"Yu-Hsiang",
""
],
[
"Hsu",
"Po-Chun",
""
],
[
"Shiu",
"Da-shan",
""
]
] |
2405.16461 | Mathew D. Penrose | Mathew D. Penrose and Xiaochuan Yang | On the rate of convergence in the Hall-Janson coverage theorem | 24 pages, one figure. Added some extra explanations and a diagram | null | null | null | math.PR | http://creativecommons.org/licenses/by/4.0/ | Consider a spherical Poisson Boolean model $Z$ in Euclidean $d$-space with $d \geq 2$, with Poisson intensity $t$ and radii distributed like $rY$ with $r \geq 0$ a scaling parameter and $Y$ a fixed nonnegative random variable with finite $(2d-2)$-nd moment (or if $d=2$, a finite $(2 + \varepsilon)$-moment condition for some $\varepsilon >0$). Let $A \subset {\bf R}^d$ be compact with a nice boundary. Let $\alpha $ be the expected volume of a ball of radius $Y$, and suppose $r=r(t)$ is chosen so that $\alpha t r^d - \log t - (d-1) \log \log t$ is a constant independent of $t$. A classical result of Hall and of Janson determines the (non-trivial) large-$t$ limit of the probability that $A$ is fully covered by $Z$. In this paper we provide an $O((\log \log t)/\log t)$ bound on the rate of convergence in that result. With a slight adjustment to $r(t)$, this can be improved to $O(1/\log t)$. | [
{
"version": "v1",
"created": "Sun, 26 May 2024 07:10:00 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Jun 2024 10:15:37 GMT"
},
{
"version": "v3",
"created": "Wed, 26 Jun 2024 10:23:42 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 16:44:04 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Penrose",
"Mathew D.",
""
],
[
"Yang",
"Xiaochuan",
""
]
] |
2405.16680 | Taewan Kim | Taewan Kim, Abhinav G. Kamath, Niyousha Rahimi, Jasper Corleis, Beh\c{c}et A\c{c}{\i}kme\c{s}e, Mehran Mesbahi | Six-Degree-of-Freedom Aircraft Landing Trajectory Planning with Runway Alignment | This article has been accepted to JGCD | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a numerical optimization algorithm for generating approach and landing trajectories for a six-degree-of-freedom (6-DoF) aircraft. We improve on the existing research on aircraft landing trajectory generation by formulating the trajectory optimization problem with additional real-world operational constraints, including 6-DoF aircraft dynamics, runway alignment, constant wind field, and obstacle avoidance, to obtain a continuous-time nonconvex optimal control problem. Particularly, the runway alignment constraint enforces the trajectory of the aircraft to be aligned with the runway only during the final approach phase. This is a novel feature that is essential for preventing an approach that is either too steep or too shallow. The proposed method models the runway alignment constraint through a multi-phase trajectory planning scheme, imposing alignment conditions exclusively during the final approach phase. We compare this formulation with the existing state-triggered constraint formulation for runway alignment. To solve the formulated problem, we design a novel sequential convex programming algorithm called xPTR that extends the penalized trust-region (PTR) algorithm by incorporating an extrapolation step to expedite convergence. We validate the proposed method through extensive numerical simulations, including a Monte Carlo study, to evaluate the robustness of the algorithm to varying initial conditions. | [
{
"version": "v1",
"created": "Sun, 26 May 2024 20:14:37 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 21:10:06 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kim",
"Taewan",
""
],
[
"Kamath",
"Abhinav G.",
""
],
[
"Rahimi",
"Niyousha",
""
],
[
"Corleis",
"Jasper",
""
],
[
"Açıkmeşe",
"Behçet",
""
],
[
"Mesbahi",
"Mehran",
""
]
] |
2405.18410 | Mahrokh Najaf | Mahrokh Najaf, Gregory Ongie | Towards a Sampling Theory for Implicit Neural Representations | IEEE Asilomar 2024 | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Implicit neural representations (INRs) have emerged as a powerful tool for solving inverse problems in computer vision and computational imaging. INRs represent images as continuous domain functions realized by a neural network taking spatial coordinates as inputs. However, unlike traditional pixel representations, little is known about the sample complexity of estimating images using INRs in the context of linear inverse problems. Towards this end, we study the sampling requirements for recovery of a continuous domain image from its low-pass Fourier coefficients by fitting a single hidden-layer INR with ReLU activation and a Fourier features layer using a generalized form of weight decay regularization. Our key insight is to relate minimizers of this non-convex parameter space optimization problem to minimizers of a convex penalty defined over an infinite-dimensional space of measures. We identify a sufficient number of samples for which an image realized by a width-1 INR is exactly recoverable by solving the INR training problem, and give a conjecture for the general width-$W$ case. To validate our theory, we empirically assess the probability of achieving exact recovery of images realized by low-width single hidden-layer INRs, and illustrate the performance of INR on super-resolution recovery of more realistic continuous domain phantom images. | [
{
"version": "v1",
"created": "Tue, 28 May 2024 17:53:47 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 02:23:11 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Najaf",
"Mahrokh",
""
],
[
"Ongie",
"Gregory",
""
]
] |
2405.19599 | Nandini Ananth | Elliot C. Eklund, Nandini Ananth | Hybrid Quantum Algorithm for Simulating Real-Time Thermal Correlation Functions | null | null | null | null | quant-ph physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a hybrid Path Integral Monte Carlo (hPIMC) algorithm to calculate real-time quantum thermal correlation functions and demonstrate its application to open quantum systems. The hPIMC algorithm leverages the successes of classical PIMC as a computational tool for high-dimensional system studies by exactly simulating dissipation using the Feynman-Vernon influence functional on a classical computer. We achieve a quantum speed-up over the classical algorithm by computing short-time matrix elements of the quantum propagator on a quantum computer. We show that the component of imaginary-time evolution can be performed accurately using the recently developed Probabilistic Imaginary-Time Evolution (PITE) algorithm, and we introduce a novel low-depth circuit for approximate real-time evolution under the kinetic energy operator using a Discrete Variable Representation (DVR). We test the accuracy of the approximation by computing the position-position thermal correlation function of a proton transfer reaction. | [
{
"version": "v1",
"created": "Thu, 30 May 2024 01:30:13 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 15:04:50 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Eklund",
"Elliot C.",
""
],
[
"Ananth",
"Nandini",
""
]
] |
2405.20421 | Qianqi Yan | Qianqi Yan, Xuehai He, Xiang Yue, Xin Eric Wang | Worse than Random? An Embarrassingly Simple Probing Evaluation of Large Multimodal Models in Medical VQA | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Large Multimodal Models (LMMs) have shown remarkable progress in medical Visual Question Answering (Med-VQA), achieving high accuracy on existing benchmarks. However, their reliability under robust evaluation is questionable. This study reveals that when subjected to simple probing evaluation, state-of-the-art models perform worse than random guessing on medical diagnosis questions. To address this critical evaluation problem, we introduce the Probing Evaluation for Medical Diagnosis (ProbMed) dataset to rigorously assess LMM performance in medical imaging through probing evaluation and procedural diagnosis. Particularly, probing evaluation features pairing original questions with negation questions with hallucinated attributes, while procedural diagnosis requires reasoning across various diagnostic dimensions for each image, including modality recognition, organ identification, clinical findings, abnormalities, and positional grounding. Our evaluation reveals that top-performing models like GPT-4o, GPT-4V, and Gemini Pro perform worse than random guessing on specialized diagnostic questions, indicating significant limitations in handling fine-grained medical inquiries. Besides, models like LLaVA-Med struggle even with more general questions, and results from CheXagent demonstrate the transferability of expertise across different modalities of the same organ, showing that specialized domain knowledge is still crucial for improving performance. This study underscores the urgent need for more robust evaluation to ensure the reliability of LMMs in critical fields like medical diagnosis, and current LMMs are still far from applicable to those fields. | [
{
"version": "v1",
"created": "Thu, 30 May 2024 18:56:01 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Jun 2024 09:32:19 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Sep 2024 22:37:19 GMT"
},
{
"version": "v4",
"created": "Sat, 05 Oct 2024 00:09:21 GMT"
},
{
"version": "v5",
"created": "Tue, 10 Jun 2025 19:06:46 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Yan",
"Qianqi",
""
],
[
"He",
"Xuehai",
""
],
[
"Yue",
"Xiang",
""
],
[
"Wang",
"Xin Eric",
""
]
] |
2405.20525 | Elijah Pelofske | Kyle Henke, Elijah Pelofske, Garrett Kenyon, Georg Hahn | Comparing Quantum Annealing and Spiking Neuromorphic Computing for Sampling Binary Sparse Coding QUBO Problems | null | npj Unconventional Computing, 2, 13 (2025) | 10.1038/s44335-025-00028-2 | LA-UR-23-26361 | cs.ET cs.CV cs.DM cs.NE quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of computing a sparse binary representation of an image. To be precise, given an image and an overcomplete, non-orthonormal basis, we aim to find a sparse binary vector indicating the minimal set of basis vectors that when added together best reconstruct the given input. We formulate this problem with an $L_2$ loss on the reconstruction error, and an $L_0$ (or, equivalently, an $L_1$) loss on the binary vector enforcing sparsity. This yields a quadratic unconstrained binary optimization problem (QUBO), whose optimal solution(s) in general is NP-hard to find. The contribution of this work is twofold. First, we solve the sparse representation QUBOs by solving them both on a D-Wave quantum annealer with Pegasus chip connectivity via minor embedding, as well as on the Intel Loihi 2 spiking neuromorphic processor using a stochastic Non-equilibrium Boltzmann Machine (NEBM). Second, we deploy Quantum Evolution Monte Carlo with Reverse Annealing and iterated warm starting on Loihi 2 to evolve the solution quality from the respective machines. The solutions are benchmarked against simulated annealing, a classical heuristic, and the optimal solutions are computed using CPLEX. Iterated reverse quantum annealing performs similarly to simulated annealing, although simulated annealing is always able to sample the optimal solution whereas quantum annealing was not always able to. The Loihi 2 solutions that are sampled are on average more sparse than the solutions from any of the other methods. We demonstrate that both quantum annealing and neuromorphic computing are suitable for binary sparse coding QUBOs, and that Loihi 2 outperforms a D-Wave quantum annealer standard linear-schedule anneal, while iterated reverse quantum annealing performs much better than both unmodified linear-schedule quantum annealing and iterated warm starting on Loihi 2. | [
{
"version": "v1",
"created": "Thu, 30 May 2024 22:56:15 GMT"
},
{
"version": "v2",
"created": "Wed, 14 May 2025 04:37:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Henke",
"Kyle",
""
],
[
"Pelofske",
"Elijah",
""
],
[
"Kenyon",
"Garrett",
""
],
[
"Hahn",
"Georg",
""
]
] |
2405.20761 | Aditya Shankar | Aditya Shankar, J\'er\'emie Decouchant, Dimitra Gkorou, Rihan Hai, Lydia Y. Chen | Share Secrets for Privacy: Confidential Forecasting with Vertical Federated Learning | Accepted at the 20th International Conference on Availability, Reliability and Security (ARES 2025) | null | null | null | cs.LG cs.CR cs.DC | http://creativecommons.org/licenses/by/4.0/ | Vertical federated learning (VFL) is a promising area for time series forecasting in many applications, such as healthcare and manufacturing. Critical challenges to address include data privacy and over-fitting on small and noisy datasets during both training and inference. Additionally, such forecasting models must scale well with the number of parties while ensuring strong convergence and low-tuning complexity. We address these challenges and propose ``Secret-shared Time Series Forecasting with VFL'' (STV), a novel framework with the following key features: i) a privacy-preserving algorithm for forecasting with SARIMAX and autoregressive trees on vertically-partitioned data; ii) decentralised forecasting using secret sharing and multi-party computation; and iii) novel N-party algorithms for matrix multiplication and inverse operations for exact parameter optimization, giving strong convergence with minimal tuning complexity. We evaluate on six representative datasets from public and industry-specific contexts. Results demonstrate that STV's forecasting accuracy is comparable to those of centralized approaches. Our exact optimization outperforms centralized methods, including state-of-the-art diffusion models and long-short-term memory, by 23.81% on forecasting accuracy. We also evaluate scalability by examining the communication costs of exact and iterative optimization to navigate the choice between the two. STV's code and supplementary material is available online: https://github.com/adis98/STV. | [
{
"version": "v1",
"created": "Fri, 31 May 2024 12:27:38 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 13:08:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Shankar",
"Aditya",
""
],
[
"Decouchant",
"Jérémie",
""
],
[
"Gkorou",
"Dimitra",
""
],
[
"Hai",
"Rihan",
""
],
[
"Chen",
"Lydia Y.",
""
]
] |
2406.00467 | Peter Hunana | P. Hunana | Generalized Fluid Models of the Braginskii Type. Part 2. The Boltzmann Operator | Accepted to ApJ Supplement Series, 28/01/2025. Few misprints and misspellings found during the production are corrected, such as one bracket in eq. (85) and two brackets \} at the bottom of p. 79 | null | null | null | physics.plasm-ph astro-ph.SR nucl-th | http://creativecommons.org/licenses/by/4.0/ | In our previous paper (Hunana et al. 2022) we have employed the Landau collisional operator together with the moment method of Grad and considered various generalizations of the Braginskii model, such as a multi-fluid formulation of the 21- and 22-moment models valid for general masses and temperatures, where all of the considered moments are described by their evolution equations (with fully non-linear left-hand-sides). Here we consider the same models, however, we employ the Boltzmann operator and calculate the collisional contributions via expressing them through the Chapman-Cowling collisional integrals. These ``integrals'' just represent a useful mathematical technique/notation introduced roughly 100 years ago, which (in the usual semi-linear approximation) allows one to postpone specifying the particular collisional process and finish all of the calculations with the Boltzmann operator. We thus consider multi-fluid 21- and 22-moment models which are valid for a large class of elastic collisional processes describable by the Boltzmann operator. Reduction into the 13-moment approximation recovers the models of Schunk and Burgers. We only focus on the particular cases of hard spheres, Coulomb collisions, purely repulsive inverse power force $|K|/r^\nu$ and attractive force $-|K|/r^\nu$ with repulsive rigid core (or potential $V(r)=\delta(r)- |c|/r^n$, so that the particles bounce from each other when they meet), but other cases can be found in the literature. In the Appendix, we introduce the Boltzmann operator in a way suitable for newcomers and we discuss a surprisingly simple recipe how to calculate the collisional contributions with analytic software. | [
{
"version": "v1",
"created": "Sat, 01 Jun 2024 15:31:15 GMT"
},
{
"version": "v2",
"created": "Tue, 03 Dec 2024 11:28:27 GMT"
},
{
"version": "v3",
"created": "Tue, 28 Jan 2025 12:28:47 GMT"
},
{
"version": "v4",
"created": "Tue, 04 Feb 2025 17:35:50 GMT"
},
{
"version": "v5",
"created": "Fri, 07 Feb 2025 18:06:19 GMT"
},
{
"version": "v6",
"created": "Wed, 11 Jun 2025 11:23:19 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Hunana",
"P.",
""
]
] |
2406.01078 | Han Sun | Han Sun, Yunkang Cao, Hao Dong, Olga Fink | Unseen Visual Anomaly Generation | 8 pages excluding supplementary | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Visual anomaly detection (AD) presents significant challenges due to the scarcity of anomalous data samples. While numerous works have been proposed to synthesize anomalous samples, these synthetic anomalies often lack authenticity or require extensive training data, limiting their applicability in real-world scenarios. In this work, we propose Anomaly Anything (AnomalyAny), a novel framework that leverages Stable Diffusion (SD)'s image generation capabilities to generate diverse and realistic unseen anomalies. By conditioning on a single normal sample during test time, AnomalyAny is able to generate unseen anomalies for arbitrary object types with text descriptions. Within AnomalyAny, we propose attention-guided anomaly optimization to direct SD attention on generating hard anomaly concepts. Additionally, we introduce prompt-guided anomaly refinement, incorporating detailed descriptions to further improve the generation quality. Extensive experiments on MVTec AD and VisA datasets demonstrate AnomalyAny's ability in generating high-quality unseen anomalies and its effectiveness in enhancing downstream AD performance. | [
{
"version": "v1",
"created": "Mon, 03 Jun 2024 07:58:09 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Dec 2024 12:00:36 GMT"
},
{
"version": "v3",
"created": "Mon, 19 May 2025 15:41:06 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 15:03:25 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Sun",
"Han",
""
],
[
"Cao",
"Yunkang",
""
],
[
"Dong",
"Hao",
""
],
[
"Fink",
"Olga",
""
]
] |
2406.01237 | Debamita Chatterjee | Debamita Chatterjee, Riddhi Shah | Characterisation of distal actions of automorphisms on the space of one-parameter subgroups of Lie groups | The typos have been fixed. The text have been rearranged for better readability. An example is added for better understanding | null | null | null | math.DS math.GR | http://creativecommons.org/licenses/by/4.0/ | For a connected Lie group $G$ and an automorphism $T$ of $G$, we consider the action of $T$ on Sub$_G$, the compact space of closed subgroups of $G$ endowed with the Chabauty topology. We study the action of $T$ on Sub$^p_G$, the closure in Sub$_G$ of the set of closed one-parameter subgroups of $G$. We relate the distality of the $T$-action on Sub$^p_G$ with that of the $T$-action on $G$ and characterise the same in terms of compactness of the closed subgroup generated by $T$ in Aut$(G)$ when $T$ acts distally on the maximal central torus and $G$ is not a vector group. We extend these results to the action of a subgroup of Aut$(G)$, and equate the distal action of any closed subgroup ${\mathcal H}$ on Sub$^p_G$ with that of every element in ${\mathcal H}$. Moreover, we show that a connected Lie group $G$ acts distally on Sub$^p_G$ by conjugation if and only if $G$ is either compact or it is isomorphic to a direct product of a compact group and a vector group. Some of our results extend those of Shah and Yadav. | [
{
"version": "v1",
"created": "Mon, 03 Jun 2024 11:56:02 GMT"
},
{
"version": "v2",
"created": "Sat, 30 Nov 2024 16:19:12 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:43:42 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chatterjee",
"Debamita",
""
],
[
"Shah",
"Riddhi",
""
]
] |
2406.02094 | Daniel Gaina | Guillermo Badia and Daniel Gaina and Alexander Knapp and Tomasz Kowalski and Martin Wirsing | Hybrid-Dynamic Ehrenfeucht-Fraisse Games | null | null | null | null | cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ehrenfeucht-Fraisse games provide means to characterize elementary equivalence for first-order logic, and by standard translation also for modal logics. We propose a novel generalization of Ehrenfeucht- Fraisse games to hybrid-dynamic logics which is direct and fully modular: parameterized by the features of the hybrid language we wish to include, for instance, the modal and hybrid language operators as well as first-order existential quantification. We use these games to establish a new modular Fraisse-Hintikka Theorem for hybrid-dynamic propositional logic and its various fragments. We study the relationship between countable game equivalence (determined by countable Ehrenfeucht- Fraisse games) and bisimulation (determined by countable back-and-forth systems). In general, the former turns out to be weaker than the latter, but under certain conditions on the language, the two coincide. We also use games to prove that for reachable image-finite Kripke structures elementary equivalence implies isomorphism. | [
{
"version": "v1",
"created": "Tue, 04 Jun 2024 08:26:11 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Dec 2024 06:05:34 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:34:49 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Badia",
"Guillermo",
""
],
[
"Gaina",
"Daniel",
""
],
[
"Knapp",
"Alexander",
""
],
[
"Kowalski",
"Tomasz",
""
],
[
"Wirsing",
"Martin",
""
]
] |
2406.02307 | Sebastian Gartzke | Sebastian Gartzke, Shanshan Wang, Thomas Guhr, Michael Schreckenberg | Correlations in Motion: A Simple Response-Based Analysis of Traffic Flow | null | null | null | null | physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Why does a traffic jams form out of nowhere, and why does it stretch for kilometers even after the initial cause is passed? This study examines how congestion moves and spreads across motorways using a surprisingly simple method: response functions. These functions are based purely on data and show how changes in traffic flow, density, and velocity are connected over time and space. Using real-world data from German motorways, we track how traffic reacts to earlier disturbances, capturing the waves of slowing and accelerating that drivers experience in stop-and-go traffic. The results demonstrate how congestion propagates and how its rhythm can be measured and predicted. Unlike complex traffic models, this approach requires no simulations or assumptions about driver behavior. It works directly from the information provided by the road. The goal is clear: to understand congestion better so that we can manage it more effectively and perhaps spend less time stuck in it. | [
{
"version": "v1",
"created": "Tue, 04 Jun 2024 13:38:45 GMT"
},
{
"version": "v2",
"created": "Thu, 28 Nov 2024 10:13:11 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 12:17:09 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Gartzke",
"Sebastian",
""
],
[
"Wang",
"Shanshan",
""
],
[
"Guhr",
"Thomas",
""
],
[
"Schreckenberg",
"Michael",
""
]
] |
2406.02719 | Pratik Manwani | P. Manwani, Y. Kang, J. Mann, B. Naranjo, G. Andonian, J. B. Rosenzweig (Department of Physics and Astronomy, UCLA, Los Angeles, California, USA) | Analysis of the blowout plasma wakefields produced by drive beams with elliptical symmetry | 8 pages, 4 figures | null | null | null | physics.acc-ph physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the underdense (blowout) regime of plasma wakefield acceleration (PWFA), the particle beam is denser than the plasma. Under these conditions, the plasma electrons are nearly completely rarefacted from the beam channel, resulting in a nominally uniform ion column. Extensive investigations of this interaction assuming axisymmetry have been undertaken. However, the plasma blowout produced by a transversely asymmetric driver possesses quite different characteristics. They create an asymmetric plasma rarefaction region (bubble) which leads to asymmetric focusing in the two transverse planes. This is also accompanied by an undesired non-uniform accelerating gradient. The asymmetric blowout cross-section is found through simulation to be elliptical, and treating it as such permits a simple extension of the symmetric theory. In particular, focusing fields linear in both transverse directions exist in the bubble. The form of the wake potential and the concomitant matching conditions in this elliptical cavity are discussed in this paper. We also discuss bubble boundary estimation in the long driver limit and applications of the asymmetric features of the wakefield. | [
{
"version": "v1",
"created": "Tue, 04 Jun 2024 19:03:13 GMT"
},
{
"version": "v2",
"created": "Sat, 09 Nov 2024 23:40:07 GMT"
},
{
"version": "v3",
"created": "Wed, 20 Nov 2024 06:36:05 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 21:20:35 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Manwani",
"P.",
"",
"Department of Physics and Astronomy, UCLA, Los Angeles, California, USA"
],
[
"Kang",
"Y.",
"",
"Department of Physics and Astronomy, UCLA, Los Angeles, California, USA"
],
[
"Mann",
"J.",
"",
"Department of Physics and Astronomy, UCLA, Los Angeles, California, USA"
],
[
"Naranjo",
"B.",
"",
"Department of Physics and Astronomy, UCLA, Los Angeles, California, USA"
],
[
"Andonian",
"G.",
"",
"Department of Physics and Astronomy, UCLA, Los Angeles, California, USA"
],
[
"Rosenzweig",
"J. B.",
"",
"Department of Physics and Astronomy, UCLA, Los Angeles, California, USA"
]
] |
2406.03724 | Arvin Lamando | Malte Gerhold, Arvin Lamando, Franz Luef | Linear Deformations of Heisenberg Modules and Gabor Frames | Fixed typos. Added a section on generalized Fell's condition | null | null | null | math.OA math.FA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Heisenberg modules over noncommutative tori may also be viewed as Gabor frames. Building on this fact, we relate to deformations of noncommutative tori a bundle of Banach spaces induced by Heisenberg modules. The construction of this bundle of Banach spaces rests on deformation results of Gabor frames with windows in Feichtinger's algebra due to Feichtinger and Kaiblinger. We extend some of these results to Heisenberg modules, \eg we establish an analog of the results by Feichtinger-Kaiblinger and a Balian-Low theorem. Finally, we extend our results to several generators on the bundle of Heisenberg modules and show that they provide a generalized Fell's condition on the bundle of noncommutative tori. | [
{
"version": "v1",
"created": "Thu, 06 Jun 2024 03:45:54 GMT"
},
{
"version": "v2",
"created": "Fri, 07 Jun 2024 01:47:03 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:33:15 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Gerhold",
"Malte",
""
],
[
"Lamando",
"Arvin",
""
],
[
"Luef",
"Franz",
""
]
] |
2406.04126 | Davor Dragicevic | Lucas Backes and Davor Dragicevic | A characterization of $(\mu,\nu)$-dichotomies via admissibility | Revised version. Accepted for publication in Mathematische Nachrichten | null | null | null | math.DS math.CA | http://creativecommons.org/licenses/by/4.0/ | We present a characterization of $(\mu,\nu)$-dichotomies in terms of the admissibility of certain pairs of weighted spaces for nonautonomous discrete time dynamics acting on Banach spaces. Our general framework enables us to treat various settings in which no similar result has been previously obtained as well as to recover and refine several known results. We emphasize that our results hold without any bounded growth assumption and the statements make no use of Lyapunov norms. Moreover, as a consequence of our characterization, we study the robustness of $(\mu, \nu)$-dichotomies, i.e. we show that this notion persists under small but very general linear perturbations. | [
{
"version": "v1",
"created": "Thu, 06 Jun 2024 14:48:14 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 09:53:56 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Backes",
"Lucas",
""
],
[
"Dragicevic",
"Davor",
""
]
] |
2406.04968 | Isabella Danhoni | Isabella Danhoni, Jordi Salinas San Martin, Jacquelyn Noronha-Hostler | Shear viscosity from perturbative Quantum Chromodynamics to the hadron resonance gas at finite baryon, strangeness, and electric charge densities | null | null | 10.1103/gckr-c2yg | null | hep-ph nucl-th | http://creativecommons.org/licenses/by/4.0/ | Through model-to-data comparisons from heavy-ion collisions, it has been shown that the Quark Gluon Plasma has an extremely small shear viscosity at vanishing densities. At large baryon densities, significantly less is known about the nature of the shear viscosity from Quantum Chromodynamics (QCD). Within heavy-ion collisions, there are three conserved charges: baryon number (B), strangeness (S), and electric charge (Q). Here we calculate the shear viscosity in two limits using perturbative QCD and an excluded-volume hadron resonance gas at finite BSQ densities. We then develop a framework that interpolates between these two limits such that shear viscosity is possible to calculate across a wide range of finite BSQ densities. We find that the pQCD and hadron resonance gas calculations have different BSQ densities dependence such that a rather non-trivial shear viscosity appears at finite densities. | [
{
"version": "v1",
"created": "Fri, 07 Jun 2024 14:33:17 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 19:30:36 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Danhoni",
"Isabella",
""
],
[
"Martin",
"Jordi Salinas San",
""
],
[
"Noronha-Hostler",
"Jacquelyn",
""
]
] |
2406.06144 | Jiaming Ji | Jiaming Ji, Kaile Wang, Tianyi Qiu, Boyuan Chen, Jiayi Zhou, Changye Li, Hantao Lou, Juntao Dai, Yunhuai Liu, Yaodong Yang | Language Models Resist Alignment: Evidence From Data Compression | Accepted by ACL2025 Main | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models (LLMs) may exhibit unintended or undesirable behaviors. Recent works have concentrated on aligning LLMs to mitigate harmful outputs. Despite these efforts, some anomalies indicate that even a well-conducted alignment process can be easily circumvented, whether intentionally or accidentally. Does alignment fine-tuning yield have robust effects on models, or are its impacts merely superficial? In this work, we make the first exploration of this phenomenon from both theoretical and empirical perspectives. Empirically, we demonstrate the $\mathbf{elasticity}$ of post-alignment models, i.e., the tendency to revert to the behavior distribution formed during the pre-training phase upon further fine-tuning. Leveraging compression theory, we formally deduce that fine-tuning disproportionately undermines alignment relative to pre-training, potentially by orders of magnitude. We validate the presence of elasticity through experiments on models of varying types and scales. Specifically, we find that model performance declines rapidly before reverting to the pre-training distribution, after which the rate of decline drops significantly. Furthermore, we further reveal that elasticity positively correlates with the increased model size and the expansion of pre-training data. Our findings underscore the need to address the inherent elasticity of LLMs to mitigate their resistance to alignment. The model weight and code are available at pku-lm-resist-alignment.github.io. | [
{
"version": "v1",
"created": "Mon, 10 Jun 2024 10:03:16 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Jun 2024 06:46:14 GMT"
},
{
"version": "v3",
"created": "Fri, 20 Dec 2024 16:25:12 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 17:23:47 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ji",
"Jiaming",
""
],
[
"Wang",
"Kaile",
""
],
[
"Qiu",
"Tianyi",
""
],
[
"Chen",
"Boyuan",
""
],
[
"Zhou",
"Jiayi",
""
],
[
"Li",
"Changye",
""
],
[
"Lou",
"Hantao",
""
],
[
"Dai",
"Juntao",
""
],
[
"Liu",
"Yunhuai",
""
],
[
"Yang",
"Yaodong",
""
]
] |
2406.08726 | Eve Fleisig | Genevieve Smith, Eve Fleisig, Madeline Bossi, Ishita Rustagi, Xavier Yin | Standard Language Ideology in AI-Generated Language | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Standard language ideology is reflected and reinforced in language generated by large language models (LLMs). We present a faceted taxonomy of open problems that illustrate how standard language ideology manifests in AI-generated language, alongside implications for minoritized language communities and society more broadly. We introduce the concept of standard AI-generated language ideology, a process through which LLMs position "standard" languages--particularly Standard American English (SAE)--as the linguistic default, reinforcing the perception that SAE is the most "appropriate" language. We then discuss ongoing tensions around what constitutes desirable system behavior, as well as advantages and drawbacks of generative AI tools attempting, or refusing, to imitate different English language varieties. Rather than prescribing narrow technical fixes, we offer three recommendations for researchers, practitioners, and funders that focus on shifting structural conditions and supporting more emancipatory outcomes for diverse language communities. | [
{
"version": "v1",
"created": "Thu, 13 Jun 2024 01:08:40 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 16:54:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Smith",
"Genevieve",
""
],
[
"Fleisig",
"Eve",
""
],
[
"Bossi",
"Madeline",
""
],
[
"Rustagi",
"Ishita",
""
],
[
"Yin",
"Xavier",
""
]
] |
2406.09113 | Dorota Skowron | Dorota M. Skowron, Ronald Drimmel, Shourya Khanna, Alessandro Spagna, Eloisa Poggio, Pau Ramos | The Milky Way as Seen by Classical Cepheids I: Distances Based on Mid-infrared Photometry | 18 pages, 24 figures, 3 tables; published in to ApJS, The catalog of Cepheid distances is made available online | 2025, ApJS, 278, 57 | 10.3847/1538-4365/adc3f3 | null | astro-ph.GA astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Classical Cepheids are the archetype of the standard candle, thanks to the period-luminosity relation which allows to measure their intrinsic brightness. They are also relatively young and bright, potentially making them excellent tracers of the young stellar population that is responsible for shaping the visible aspect of our Galaxy, the Milky Way. However, being observers embedded in the dusty interstellar medium of the Galaxy, deriving reliable photometric distances to classical Cepheids of the Milky Way is a challenge. The typical approach is to use "reddening-free" indices, such as Wesenheit magnitudes, to obviate the need for an extinction correction. However, this approach could lead to unknown systematics - especially toward the inner Galaxy - as its assumption of a universal total-to-selective extinction ratio is not satisfied, particularly in lines of sight where the extinction is high and crosses spiral arms. We instead estimate new distances for 3424 Cepheids based on mid-IR photometry from WISE, which suffers minimally from extinction, and by adopting a 3D extinction map to calculate the necessary (albeit small) extinction corrections. We show that our distances are consistent with Gaia's parallaxes for the subset with relative parallax errors smaller than 10%, verifying that our mean distance errors are of the order of 6% and that the mean parallax zero point for this subsample is 7 $\mu$as. | [
{
"version": "v1",
"created": "Thu, 13 Jun 2024 13:43:16 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 10:27:00 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Skowron",
"Dorota M.",
""
],
[
"Drimmel",
"Ronald",
""
],
[
"Khanna",
"Shourya",
""
],
[
"Spagna",
"Alessandro",
""
],
[
"Poggio",
"Eloisa",
""
],
[
"Ramos",
"Pau",
""
]
] |
2406.09568 | Ohad Lib | Edward Tananyan, Ohad Lib, Michal Zimmerman, and Yaron Bromberg | Reciprocity-assisted ghost imaging through dynamic random media | null | null | 10.1364/OL.542962 | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | Ghost imaging enables the imaging of an object using intensity correlations between a single-pixel detector placed behind the object and a camera that records light that did not interact with the object. The object and the camera are often placed at conjugate planes to ensure correlated illumination patterns. Here, we show how the combined effect of optical reciprocity and the memory effect in a random medium gives rise to correlations between two beams that traverse the random medium in opposite directions. Using these correlations, we demonstrate a ghost imaging scheme in which the object and camera are placed at opposite ends of the random medium and illuminated by counter-propagating beams that can potentially be emitted by two different sources. | [
{
"version": "v1",
"created": "Thu, 13 Jun 2024 20:18:52 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tananyan",
"Edward",
""
],
[
"Lib",
"Ohad",
""
],
[
"Zimmerman",
"Michal",
""
],
[
"Bromberg",
"Yaron",
""
]
] |
2406.09956 | Lina Vandr\'e | Lina Vandr\'e, Jarn de Jong, Frederik Hahn, Adam Burchardt, Otfried G\"uhne, Anna Pappa | Distinguishing Graph States by the Properties of Their Marginals | 14 pages of main text + 5 pages of appendices, 11+4 figures, 3 tables | Phys. Rev. A 111, 052449 (2025) | 10.1103/PhysRevA.111.052449 | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph states are a class of multi-partite entangled quantum states that are ubiquitous in quantum information. We study equivalence relations between graph states under local unitaries (LU) to obtain distinguishing methods both in local and in networked settings. Based on the marginal structure of graph states, we introduce a family of easy-to-compute LU-invariants. We show that these invariants uniquely identify the entanglement classes of every graph state up to 8 qubits and discuss their reliability for larger numbers of qubits. To handle larger graphs, we generalize tools to test for local Clifford (LC) equivalence of graph states that work by condensing large graphs into smaller graphs. In turn, we show that statements on the equivalence of these smaller graphs (which are easier to compute) can be used to infer statements on the equivalence of the original, larger graphs. We analyze LU-equivalence in two key settings - with and without allowing for the permutation of qubits. We identify entanglement classes, whose marginal structure does not allow us to distinguish them. As a result, we increase the bound on the number of qubits where the LU-LC conjecture holds from 8 to 10 qubits in the setting where qubit permutations are allowed. | [
{
"version": "v1",
"created": "Fri, 14 Jun 2024 12:03:10 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:12:01 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Vandré",
"Lina",
""
],
[
"de Jong",
"Jarn",
""
],
[
"Hahn",
"Frederik",
""
],
[
"Burchardt",
"Adam",
""
],
[
"Gühne",
"Otfried",
""
],
[
"Pappa",
"Anna",
""
]
] |
2406.10322 | Sophie Ostmeier | Sophie Ostmeier, Brian Axelrod, Maya Varma, Michael E. Moseley, Akshay Chaudhari, Curtis Langlotz | LieRE: Lie Rotational Positional Encodings | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transformer architectures depend on explicit position encodings to capture token positional information. Rotary Position Encoding (RoPE) has emerged as a popular choice in language models due to its efficient encoding of relative position information through key-query rotations. However, RoPE faces significant limitations beyond language processing: it is constrained to one-dimensional sequence data and, even with learnable phases, offers limited representational capacity. We address these challenges with Lie Relative Encodings (LieRE), which generalizes RoPE to high-dimensional rotation matrices by leveraging their Lie group structure. Through extensive evaluation on three image datasets across 2D and 3D classification tasks, LieRE achieves 1.5% improvement over state-of-the-art baselines on 2D tasks and 1% on 3D tasks, while demonstrating superior generalization to higher resolutions. Our implementation is computationally efficient, with results reproducible on 4 A100 GPUs in 30 minutes on CIFAR100. Our code is available at https://github.com/StanfordMIMI/LieRE. | [
{
"version": "v1",
"created": "Fri, 14 Jun 2024 17:41:55 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Oct 2024 16:06:18 GMT"
},
{
"version": "v3",
"created": "Tue, 18 Feb 2025 16:52:29 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 20:39:51 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ostmeier",
"Sophie",
""
],
[
"Axelrod",
"Brian",
""
],
[
"Varma",
"Maya",
""
],
[
"Moseley",
"Michael E.",
""
],
[
"Chaudhari",
"Akshay",
""
],
[
"Langlotz",
"Curtis",
""
]
] |
2406.12338 | Xiulin Wang | Carla Schenker, Xiulin Wang, David Horner, Morten A. Rasmussen, Evrim Acar | PARAFAC2-based Coupled Matrix and Tensor Factorizations with Constraints | 15 pages, 15 figures,1 table | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data fusion models based on Coupled Matrix and Tensor Factorizations (CMTF) have been effective tools for joint analysis of data from multiple sources. While the vast majority of CMTF models are based on the strictly multilinear CANDECOMP/PARAFAC (CP) tensor model, recently also the more flexible PARAFAC2 model has been integrated into CMTF models. PARAFAC2 tensor models can handle irregular/ragged tensors and have shown to be especially useful for modelling dynamic data with unaligned or irregular time profiles. However, existing PARAFAC2-based CMTF models have limitations in terms of possible regularizations on the factors and/or types of coupling between datasets. To address these limitations, in this paper we introduce a flexible algorithmic framework that fits PARAFAC2-based CMTF models using Alternating Optimization (AO) and the Alternating Direction Method of Multipliers (ADMM). The proposed framework allows to impose various constraints on all modes and linear couplings to other matrix-, CP- or PARAFAC2-models. Experiments on various simulated and a real dataset demonstrate the utility and versatility of the proposed framework as well as its benefits in terms of accuracy and efficiency in comparison with state-of-the-art methods. | [
{
"version": "v1",
"created": "Tue, 18 Jun 2024 07:05:31 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 00:17:52 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Schenker",
"Carla",
""
],
[
"Wang",
"Xiulin",
""
],
[
"Horner",
"David",
""
],
[
"Rasmussen",
"Morten A.",
""
],
[
"Acar",
"Evrim",
""
]
] |
2406.12517 | Kun Xu | Yiqing Lin and Kun Xu | Propagation of chaos for mean-field reflected BSDEs with jumps | null | null | null | null | math.PR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study a class of mean-field reflected backward stochastic differential equations (MF-RBSDEs) driven by a marked point process and also analyze MF-RBSDEs driven by a Poisson process. Based on a $g$-expectation representation lemma, we give the existence and uniqueness of the particle system of MF-RBSDEs driven by a marked point process under Lipschitz generator conditions and obtain a convergence result of this system. We also establish the well-posedness of the MF-RBSDEs driven by a Poisson process and the convergence rate of the corresponding particle system towards the solution to the MF-RBSDEs driven by a Poisson process under bounded terminal, bounded obstacle conditions. | [
{
"version": "v1",
"created": "Tue, 18 Jun 2024 11:44:22 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 08:38:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lin",
"Yiqing",
""
],
[
"Xu",
"Kun",
""
]
] |
2406.14157 | David Wiltshire | Marco Galoppo, David L. Wiltshire | Exact solutions for differentially rotating galaxies in general relativity | 13 pages, 3 figures; v3: Significant error corrected, reinforcing physical interpretation. Fits to GAIA-DR3 data added | null | null | null | gr-qc astro-ph.CO astro-ph.GA hep-ph hep-th | http://creativecommons.org/licenses/by/4.0/ | Two classes of stationary axisymmetric solutions of Einstein's equations for isolated differentially rotating matter sources are presented. The asymptotic regime is extracted, with attention to quasilocal gravitational energy, shear and angular momentum related by the Raychaudhuri equation. At outer boundaries the quasilocal angular momentum and energy densities vanish, defining novel geometric structures - the vortex surface and the rotosurface. Solutions with a rotosurface present a new notion of asymptotic flatness: vorticity, shear and Kretschmann scalars vanish at a finite radial distance from the symmetry centre. Synthetic rotation curves for the Milky Way are presented, fit to GAIA-DR3 data, and potential astrophysical signatures are suggested. Whether or not abundant collisionless dark matter exists, the new solutions very strongly support suggestions that the phenomenology of galactic rotation curves be fundamentally reconsidered, for consistency with general relativity. | [
{
"version": "v1",
"created": "Thu, 20 Jun 2024 09:54:18 GMT"
},
{
"version": "v2",
"created": "Thu, 09 Jan 2025 18:41:16 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 03:55:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Galoppo",
"Marco",
""
],
[
"Wiltshire",
"David L.",
""
]
] |
2406.14230 | Han Jiang | Han Jiang, Xiaoyuan Yi, Zhihua Wei, Ziang Xiao, Shu Wang, Xing Xie | Raising the Bar: Investigating the Values of Large Language Models via Generative Evolving Testing | ICML 2025 | null | null | null | cs.CL cs.AI cs.CY | http://creativecommons.org/licenses/by/4.0/ | Warning: Contains harmful model outputs. Despite significant advancements, the propensity of Large Language Models (LLMs) to generate harmful and unethical content poses critical challenges. Measuring value alignment of LLMs becomes crucial for their regulation and responsible deployment. Although numerous benchmarks have been constructed to assess social bias, toxicity, and ethical issues in LLMs, those static benchmarks suffer from evaluation chronoeffect, in which, as models rapidly evolve, existing benchmarks may leak into training data or become saturated, overestimating ever-developing LLMs. To tackle this problem, we propose GETA, a novel generative evolving testing approach based on adaptive testing methods in measurement theory. Unlike traditional adaptive testing methods that rely on a static test item pool, GETA probes the underlying moral boundaries of LLMs by dynamically generating test items tailored to model capability. GETA co-evolves with LLMs by learning a joint distribution of item difficulty and model value conformity, thus effectively addressing evaluation chronoeffect. We evaluated various popular LLMs with GETA and demonstrated that 1) GETA can dynamically create difficulty-tailored test items and 2) GETA's evaluation results are more consistent with models' performance on unseen OOD and i.i.d. items, laying the groundwork for future evaluation paradigms. | [
{
"version": "v1",
"created": "Thu, 20 Jun 2024 11:51:00 GMT"
},
{
"version": "v2",
"created": "Fri, 12 Jul 2024 03:47:21 GMT"
},
{
"version": "v3",
"created": "Mon, 03 Feb 2025 10:33:17 GMT"
},
{
"version": "v4",
"created": "Tue, 27 May 2025 04:16:21 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 07:47:45 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Jiang",
"Han",
""
],
[
"Yi",
"Xiaoyuan",
""
],
[
"Wei",
"Zhihua",
""
],
[
"Xiao",
"Ziang",
""
],
[
"Wang",
"Shu",
""
],
[
"Xie",
"Xing",
""
]
] |
2406.14917 | Melvin Wong | Melvin Wong, Jiao Liu, Thiago Rios, Stefan Menzel, Yew Soon Ong | LLM2TEA: Agentic AI Designer Finds Innovative Objects with Generative Evolutionary Multitasking | This work has been submitted to the IEEE for review | null | null | null | cs.AI cs.CL cs.CV cs.LG cs.NE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we introduce LLM-driven MultiTask Evolutionary Algorithm (LLM2TEA), the first agentic AI designer within a generative evolutionary multitasking (GEM) framework that promotes the crossover and synergy of designs from multiple domains, leading to innovative solutions that transcend individual disciplines. Of particular interest is the discovery of objects that are not only innovative but also conform to the physical specifications of the real world in science and engineering. LLM2TEA comprises a large language model to initialize a population of genotypes (defined by text prompts) describing the objects of interest, a text-to-3D generative model to produce phenotypes from these prompts, a classifier to interpret the semantic representations of the objects, and a physics simulation model to assess their physical properties. We propose several novel LLM-based multitask evolutionary operators to guide the search toward the discovery of high-performing practical objects. Experimental results in conceptual design optimization validate the effectiveness of LLM2TEA, revealing from 97\% to 174\% improvement in the diversity of innovative objects compared to the present text-to-3D generative model baseline. In addition, more than 73\% of the generated designs have better physical performance than the top 1\% percentile of the designs generated in the baseline. Moreover, LLM2TEA generates designs that are not only aesthetically creative but also functional in real-world applications. Several of these designs have been successfully 3D-printed, emphasizing the proposed approach's capacity to transform AI-generated outputs into tangible physical objects. The designs produced by LLM2TEA meets practical requirements while showcasing creative and innovative features, underscoring its potential applications in complex design optimization and discovery. | [
{
"version": "v1",
"created": "Fri, 21 Jun 2024 07:20:51 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 13:19:51 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wong",
"Melvin",
""
],
[
"Liu",
"Jiao",
""
],
[
"Rios",
"Thiago",
""
],
[
"Menzel",
"Stefan",
""
],
[
"Ong",
"Yew Soon",
""
]
] |
2406.15287 | Nathaniel Sagman | Christian El Emam and Nathaniel Sagman | Complex affine spheres and a Bers theorem for SL(3,C) | Removed content on Beltrami equation (now contained in arXiv:2410.06175) and on harmonic maps (to be treated in a different paper) | null | null | null | math.DG math.CV math.GT | http://creativecommons.org/publicdomain/zero/1.0/ | For $S$ a closed surface of genus at least $2$, let $\mathrm{Hit}_3(S)$ be the Hitchin component of representations to $\mathrm{SL}(3,\mathbb{R}),$ equipped with the Labourie-Loftin complex structure. We construct a mapping class group equivariant holomorphic map from a large open subset of $\mathrm{Hit}_3(S)\times \overline{\mathrm{Hit}_3(S)}$ to the $\mathrm{SL}(3,\mathbb{C})$-character variety that restricts to the identity on the diagonal and to Bers' simultaneous uniformization on $\mathrm{T}(S)\times \overline{\mathrm{T}(S)}$. The open subset contains $\mathrm{Hit}_3(S)\times \overline{\mathrm{T}(S)}$ and $\mathrm{T}(S)\times \overline{\mathrm{Hit}_3(S)}$, and the image includes the holonomies of $\mathrm{SL}(3,\mathbb{C})$-opers.
The map is realized by associating pairs of Hitchin representations to immersions into $\mathbb{C}^3$ that we call complex affine spheres, which are equivalent to certain conformal harmonic maps into $\mathrm{SL}(3,\mathbb{C})/\mathrm{SO}(3,\mathbb{C})$ and to new objects called bi-Higgs bundles. Complex affine spheres are obtained by solving a second-order complex elliptic PDE that resembles both the Beltrami and Tzitz\'eica equations. To study this equation we establish analytic results that should be of independent interest. | [
{
"version": "v1",
"created": "Fri, 21 Jun 2024 16:31:49 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Jun 2024 15:56:58 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 16:51:26 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Emam",
"Christian El",
""
],
[
"Sagman",
"Nathaniel",
""
]
] |
2406.15369 | Selman Akbulut | Selman Akbulut | Corks | 1 page, no figures. A minor misprint corrected, and a historical fact inserted | null | null | null | math.HO math.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Remarks relating the various notions of corks. | [
{
"version": "v1",
"created": "Mon, 08 Apr 2024 05:46:46 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Jun 2024 07:47:30 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 12:45:10 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Akbulut",
"Selman",
""
]
] |
2406.15481 | Haneul Yoo | Haneul Yoo, Yongjin Yang, Hwaran Lee | Code-Switching Red-Teaming: LLM Evaluation for Safety and Multilingual Understanding | To appear in ACL 2025 | null | null | null | cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | As large language models (LLMs) have advanced rapidly, concerns regarding their safety have become prominent. In this paper, we discover that code-switching in red-teaming queries can effectively elicit undesirable behaviors of LLMs, which are common practices in natural language. We introduce a simple yet effective framework, CSRT, to synthesize codeswitching red-teaming queries and investigate the safety and multilingual understanding of LLMs comprehensively. Through extensive experiments with ten state-of-the-art LLMs and code-switching queries combining up to 10 languages, we demonstrate that the CSRT significantly outperforms existing multilingual red-teaming techniques, achieving 46.7% more attacks than standard attacks in English and being effective in conventional safety domains. We also examine the multilingual ability of those LLMs to generate and understand codeswitching texts. Additionally, we validate the extensibility of the CSRT by generating codeswitching attack prompts with monolingual data. We finally conduct detailed ablation studies exploring code-switching and propound unintended correlation between resource availability of languages and safety alignment in existing multilingual LLMs. | [
{
"version": "v1",
"created": "Mon, 17 Jun 2024 06:08:18 GMT"
},
{
"version": "v2",
"created": "Sat, 02 Nov 2024 06:21:44 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 05:32:09 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Yoo",
"Haneul",
""
],
[
"Yang",
"Yongjin",
""
],
[
"Lee",
"Hwaran",
""
]
] |
2406.15804 | Xianke Qiang | Xianke Qiang, Zheng Chang, Chaoxiong Ye, Timo Hamalainen, Geyong Min | Split Federated Learning Empowered Vehicular Edge Intelligence: Concept, Adaptive Design and Future Directions | null | null | 10.1109/MWC.009.2400219 | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To achieve ubiquitous intelligence in future vehicular networks, artificial intelligence (AI) is essential for extracting valuable insights from vehicular data to enhance AI-driven services. By integrating AI technologies into Vehicular Edge Computing (VEC) platforms, which provides essential storage, computing, and network resources, Vehicular Edge Intelligence (VEI) can be fully realized. Traditional centralized learning, as one of the enabling technologies for VEI, places significant strain on network bandwidth while also increasing latency and privacy concerns. Nowadays, distributed machine learning methods, such as Federated Learning (FL), Split Learning (SL), and Split Federated Learning (SFL), are widely applied in vehicular networks to support VEI. However, these methods still face significant challenges due to the mobility and constrained resources inherent in vehicular networks. In this article, we first provide an overview of the system architecture, performance metrics, and challenges associated with VEI design. Then, the adaptive design of SFL, namely Adaptive Split Federated Learning (ASFL) is introduced. The proposed ASFL scheme dynamically adapts the cut layer selection process and operates in parallel, optimizing both communication and computation efficiency while improving model performance under non-IID data distribution. Finally, we highlight future research directions to shed the light on the efficient design of SFL. | [
{
"version": "v1",
"created": "Sat, 22 Jun 2024 10:15:15 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Jun 2024 12:44:35 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 07:38:03 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Qiang",
"Xianke",
""
],
[
"Chang",
"Zheng",
""
],
[
"Ye",
"Chaoxiong",
""
],
[
"Hamalainen",
"Timo",
""
],
[
"Min",
"Geyong",
""
]
] |
2406.15853 | Hua-Lei Yin | Yu-Shuo Lu, Hua-Lei Yin, Yuan-Mei Xie, Yao Fu, Zeng-Bing Chen | Repeater-like asynchronous measurement-device-independent quantum conference key agreement | 28 pages, 9 figures | Rep. Prog. Phys. 88, 067901 (2025) | 10.1088/1361-6633/addeec | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum conference key agreement enables secure communication among multiple parties by leveraging multipartite entanglement, which is expected to play a crucial role in future quantum networks. However, its practical implementation has been severely limited by the experimental complexity and low efficiency associated with the requirement for synchronous detection of multipartite entangled states. In this work, we propose a measurement-device-independent quantum conference key agreement protocol that employs asynchronous Greenberger-Horne-Zeilinger state measurement. Our protocol enables a linear scaling of the conference key rate among multiple parties, demonstrating performance comparable to that of the single-repeater scheme in quantum networks. Additionally, we achieve intercity transmission distances with composable security under finite-key conditions. By adopting the generalized asynchronous pairing strategy, our approach eliminates the need for complex global phase locking techniques. Furthermore, by integrating asynchronous pairing with ring-interference network structure, our method provides insights for various quantum tasks beyond quantum communication, including multiparty computing and quantum repeaters. | [
{
"version": "v1",
"created": "Sat, 22 Jun 2024 13:57:21 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Mar 2025 08:57:54 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 15:47:41 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lu",
"Yu-Shuo",
""
],
[
"Yin",
"Hua-Lei",
""
],
[
"Xie",
"Yuan-Mei",
""
],
[
"Fu",
"Yao",
""
],
[
"Chen",
"Zeng-Bing",
""
]
] |
2406.16439 | Shilei Cao | Shilei Cao, Juepeng Zheng, Yan Liu, Baoquan Zhao, Ziqi Yuan, Weijia Li, Runmin Dong, Haohuan Fu | Exploring Test-Time Adaptation for Object Detection in Continually Changing Environments | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Real-world application models are commonly deployed in dynamic environments, where the target domain distribution undergoes temporal changes. Continual Test-Time Adaptation (CTTA) has recently emerged as a promising technique to gradually adapt a source-trained model to continually changing target domains. Despite recent advancements in addressing CTTA, two critical issues remain: 1) Fixed thresholds for pseudo-labeling in existing methodologies lead to low-quality pseudo-labels, as model confidence varies across categories and domains; 2) Stochastic parameter restoration methods for mitigating catastrophic forgetting fail to preserve critical information effectively, due to their intrinsic randomness. To tackle these challenges for detection models in CTTA scenarios, we present AMROD, featuring three core components. Firstly, the object-level contrastive learning module extracts object-level features for contrastive learning to refine the feature representation in the target domain. Secondly, the adaptive monitoring module dynamically skips unnecessary adaptation and updates the category-specific threshold based on predicted confidence scores to enable efficiency and improve the quality of pseudo-labels. Lastly, the adaptive randomized restoration mechanism selectively reset inactive parameters with higher possibilities, ensuring the retention of essential knowledge. We demonstrate the effectiveness of AMROD on four CTTA object detection tasks, where AMROD outperforms existing methods, especially achieving a 3.2 mAP improvement and a 20\% increase in efficiency on the Cityscapes-to-Cityscapes-C CTTA task. The code of this work is available at https://github.com/ShileiCao/AMROD. | [
{
"version": "v1",
"created": "Mon, 24 Jun 2024 08:30:03 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Jun 2024 02:16:47 GMT"
},
{
"version": "v3",
"created": "Sun, 18 Aug 2024 07:45:23 GMT"
},
{
"version": "v4",
"created": "Wed, 13 Nov 2024 09:41:00 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 07:51:33 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Cao",
"Shilei",
""
],
[
"Zheng",
"Juepeng",
""
],
[
"Liu",
"Yan",
""
],
[
"Zhao",
"Baoquan",
""
],
[
"Yuan",
"Ziqi",
""
],
[
"Li",
"Weijia",
""
],
[
"Dong",
"Runmin",
""
],
[
"Fu",
"Haohuan",
""
]
] |
2406.17761 | Shane Arora | Shane Arora, Marzena Karpinska, Hung-Ting Chen, Ipsita Bhattacharjee, Mohit Iyyer, Eunsol Choi | CaLMQA: Exploring culturally specific long-form question answering across 23 languages | 46 pages, 26 figures. Accepted as a main conference paper at ACL 2025. Code and data available at https://github.com/2015aroras/CaLMQA . Dataset expanded to 51.7K questions | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Despite rising global usage of large language models (LLMs), their ability to generate long-form answers to culturally specific questions remains unexplored in many languages. To fill this gap, we perform the first study of textual multilingual long-form QA by creating CaLMQA, a dataset of 51.7K culturally specific questions across 23 different languages. We define culturally specific questions as those that refer to concepts unique to one or a few cultures, or have different answers depending on the cultural or regional context. We obtain these questions by crawling naturally-occurring questions from community web forums in high-resource languages, and by hiring native speakers to write questions in under-resourced, rarely-studied languages such as Fijian and Kirundi. Our data collection methodologies are translation-free, enabling the collection of culturally unique questions like "Kuber iki umwami wa mbere w'uburundi yitwa Ntare?" (Kirundi; English translation: "Why was the first king of Burundi called Ntare (Lion)?"). We evaluate factuality, relevance and surface-level quality of LLM-generated long-form answers, finding that (1) for many languages, even the best models make critical surface-level errors (e.g., answering in the wrong language, repetition), especially for low-resource languages; and (2) answers to culturally specific questions contain more factual errors than answers to culturally agnostic questions -- questions that have consistent meaning and answer across many cultures. We release CaLMQA to facilitate future research in cultural and multilingual long-form QA. | [
{
"version": "v1",
"created": "Tue, 25 Jun 2024 17:45:26 GMT"
},
{
"version": "v2",
"created": "Wed, 03 Jul 2024 16:33:55 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 16:56:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Arora",
"Shane",
""
],
[
"Karpinska",
"Marzena",
""
],
[
"Chen",
"Hung-Ting",
""
],
[
"Bhattacharjee",
"Ipsita",
""
],
[
"Iyyer",
"Mohit",
""
],
[
"Choi",
"Eunsol",
""
]
] |
2406.17962 | Bohao Yang | Bohao Yang, Dong Liu, Chenghao Xiao, Kun Zhao, Chao Li, Lin Yuan, Guang Yang, Chenghua Lin | Crafting Customisable Characters with LLMs: Introducing SimsChat, a Persona-Driven Role-Playing Agent Framework | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) demonstrate remarkable ability to comprehend instructions and generate human-like text, enabling sophisticated agent simulation beyond basic behavior replication. However, the potential for creating freely customisable characters remains underexplored. We introduce the Customisable Conversation Agent Framework, which employs LLMs to simulate real-world characters through personalised characteristic feature injection, enabling diverse character creation according to user preferences. We propose the SimsConv dataset, comprising 68 customised characters and 13,971 multi-turn role-playing dialogues across 1,360 real-world scenes. Characters are initially customised using pre-defined elements (career, aspiration, traits, skills), then expanded through personal and social profiles. Building on this, we present SimsChat, a freely customisable role-playing agent incorporating various realistic settings and topic-specified character interactions. Experimental results on both SimsConv and WikiRoleEval datasets demonstrate SimsChat's superior performance in maintaining character consistency, knowledge accuracy, and appropriate question rejection compared to existing models. Our framework provides valuable insights for developing more accurate and customisable human simulacra. Our data and code are publicly available at https://github.com/Bernard-Yang/SimsChat. | [
{
"version": "v1",
"created": "Tue, 25 Jun 2024 22:44:17 GMT"
},
{
"version": "v2",
"created": "Sun, 30 Jun 2024 21:15:47 GMT"
},
{
"version": "v3",
"created": "Fri, 16 Aug 2024 08:48:26 GMT"
},
{
"version": "v4",
"created": "Thu, 16 Jan 2025 15:47:58 GMT"
},
{
"version": "v5",
"created": "Tue, 25 Feb 2025 16:30:21 GMT"
},
{
"version": "v6",
"created": "Wed, 11 Jun 2025 15:18:44 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Yang",
"Bohao",
""
],
[
"Liu",
"Dong",
""
],
[
"Xiao",
"Chenghao",
""
],
[
"Zhao",
"Kun",
""
],
[
"Li",
"Chao",
""
],
[
"Yuan",
"Lin",
""
],
[
"Yang",
"Guang",
""
],
[
"Lin",
"Chenghua",
""
]
] |
2406.18402 | Otari Sakhelashvili | Gia Dvali, Archil Kobakhidze, Otari Sakhelashvili | Hint to Supersymmetry from GR Vacuum | 20 pages | Phys.Rev.D 110 (2024) 8, 086008 | 10.1103/PhysRevD.110.086008 | null | hep-th gr-qc hep-ph | http://creativecommons.org/licenses/by/4.0/ | The $S$-matrix formulation of gravity suggests that the $\theta$-vacuum structure must not be sustained by the theory. We point out that, when applied to the vacuum of general relativity, this criterion hints to supersymmetry. The topological susceptibility of gravitational vacuum induced by Eguchi-Hanson instantons can be eliminated neither by spin-$1/2$ fermions nor by an axion coupled via them since such fermions do not provide instanton zero modes. Instead, the job is done by a spin-$3/2$ fermion, hence realizing a local supersymmetry. This scenario also necessitates the spontaneous breaking of supersymmetry and predicts the existence of axion of $R$-symmetry which gets mass exclusively from the gravitational instantons. The $R$-axion can be a viable dark matter candidate. Matching between the index and the anomaly imposes a constraint that spin-$1/2$ fermions should not contribute to the chiral gravitational anomaly. | [
{
"version": "v1",
"created": "Wed, 26 Jun 2024 14:55:18 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Dvali",
"Gia",
""
],
[
"Kobakhidze",
"Archil",
""
],
[
"Sakhelashvili",
"Otari",
""
]
] |
2406.18959 | Peng Liang | Musengamana Jean de Dieu, Peng Liang, Mojtaba Shahin, Arif Ali Khan | How Do Users Revise Architectural Related Questions on Stack Overflow: An Empirical Study | 42 pages, 7 images, 8 tables, Manuscript revision submitted to a journal (2025) | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Technical Questions and Answers (Q&A) sites, such as Stack Overflow (SO), accumulate a significant variety of information related to software development in posts from users. To ensure the quality of this information, SO encourages its users to review posts through various mechanisms (e.g., question and answer revision processes). Although Architecture Related Posts (ARPs) communicate architectural information that has a system-wide impact on development, little is known about how SO users revise information shared in ARPs. To fill this gap, we conducted an empirical study to understand how users revise Architecture Related Questions (ARQs) on SO. We manually checked 13,205 ARPs and finally identified 4,114 ARQs that contain revision information. Our main findings are that: (1) The revision of ARQs is not prevalent in SO, and an ARQ revision starts soon after this question is posted (i.e., from 1 minute onward). Moreover, the revision of an ARQ occurs before and after this question receives its first answer/architecture solution, with most revisions beginning before the first architecture solution is posted. Both Question Creators (QCs) and non-QCs actively participate in ARQ revisions, with most revisions being made by QCs. (2) A variety of information (14 categories) is missing and further provided in ARQs after being posted, among which design context, component dependency, and architecture concern are dominant information. (3) Clarify the understanding of architecture under design and improve the readability of architecture problem are the two major purposes of the further provided information in ARQs. (4) The further provided information in ARQs has several impacts on the quality of answers/architecture solutions, including making architecture solution useful, making architecture solution informative, making architecture solution relevant, among others. | [
{
"version": "v1",
"created": "Thu, 27 Jun 2024 07:42:49 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 19:08:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"de Dieu",
"Musengamana Jean",
""
],
[
"Liang",
"Peng",
""
],
[
"Shahin",
"Mojtaba",
""
],
[
"Khan",
"Arif Ali",
""
]
] |
2406.19048 | Yang Song | Yang Song, Lin Wang | BiCo-Fusion: Bidirectional Complementary LiDAR-Camera Fusion for Semantic- and Spatial-Aware 3D Object Detection | Accepted by IEEE Robotics and Automation Letters (RA-L) | IEEE Robotics and Automation Letters, Volume 10 Issue 2, 1457 - 1464, February 2025 | 10.1109/LRA.2024.3518845 | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D object detection is an important task that has been widely applied in autonomous driving. To perform this task, a new trend is to fuse multi-modal inputs, i.e., LiDAR and camera. Under such a trend, recent methods fuse these two modalities by unifying them in the same 3D space. However, during direct fusion in a unified space, the drawbacks of both modalities (LiDAR features struggle with detailed semantic information and the camera lacks accurate 3D spatial information) are also preserved, diluting semantic and spatial awareness of the final unified representation. To address the issue, this letter proposes a novel bidirectional complementary LiDAR-camera fusion framework, called BiCo-Fusion that can achieve robust semantic- and spatial-aware 3D object detection. The key insight is to fuse LiDAR and camera features in a bidirectional complementary way to enhance the semantic awareness of the LiDAR and the 3D spatial awareness of the camera. The enhanced features from both modalities are then adaptively fused to build a semantic- and spatial-aware unified representation. Specifically, we introduce Pre-Fusion consisting of a Voxel Enhancement Module (VEM) to enhance the semantic awareness of voxel features from 2D camera features and Image Enhancement Module (IEM) to enhance the 3D spatial awareness of camera features from 3D voxel features. We then introduce Unified Fusion (U-Fusion) to adaptively fuse the enhanced features from the last stage to build a unified representation. Extensive experiments demonstrate the superiority of our BiCo-Fusion against the prior arts. Project page: https://t-ys.github.io/BiCo-Fusion/. | [
{
"version": "v1",
"created": "Thu, 27 Jun 2024 09:56:38 GMT"
},
{
"version": "v2",
"created": "Sun, 01 Dec 2024 07:07:02 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 09:16:14 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Song",
"Yang",
""
],
[
"Wang",
"Lin",
""
]
] |
2406.19384 | Vedang Lad | Vedang Lad, Wes Gurnee, Max Tegmark | The Remarkable Robustness of LLMs: Stages of Inference? | For Github code see https://github.com/vdlad/Remarkable-Robustness-of-LLMs. Send all correspondence to the first author | null | null | null | cs.LG cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | We investigate the robustness of Large Language Models (LLMs) to structural interventions by deleting and swapping adjacent layers during inference. Surprisingly, models retain 72-95% of their original top-1 prediction accuracy without any fine-tuning. We find that performance degradation is not uniform across layers: interventions to the early and final layers cause the most degradation, while the model is remarkably robust to dropping middle layers. This pattern of localized sensitivity motivates our hypothesis of four stages of inference, observed across diverse model families and sizes: (1) detokenization, where local context is integrated to lift raw token embeddings into higher-level representations; (2) feature engineering, where task- and entity-specific features are iteratively refined; (3) prediction ensembling, where hidden states are aggregated into plausible next-token predictions; and (4) residual sharpening, where irrelevant features are suppressed to finalize the output distribution. Synthesizing behavioral and mechanistic evidence, we provide a framework for interpreting depth-dependent computations in LLMs. | [
{
"version": "v1",
"created": "Thu, 27 Jun 2024 17:57:03 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 16:12:50 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lad",
"Vedang",
""
],
[
"Gurnee",
"Wes",
""
],
[
"Tegmark",
"Max",
""
]
] |
2406.19683 | Xuanran Zhu | Xuanran Zhu, Chao Zhang, Zheng An, Bei Zeng | Unified Framework for Calculating Convex Roof Resource Measures | 13 pages, 5 figures, 2 tables | npj Quantum Inf 11, 56 (2025) | 10.1038/s41534-025-01012-1 | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantum resource theories (QRTs) provide a comprehensive and practical framework for the analysis of diverse quantum phenomena. A fundamental task within QRTs is the quantification of resources inherent in a given quantum state. In this letter, we introduce a unified computational framework for a class of widely utilized quantum resource measures, derived from convex roof extensions. We establish that the computation of these convex roof resource measures can be reformulated as an optimization problem over a Stiefel manifold, which can be further unconstrained through polar projection. Compared to existing methods employing semi-definite programming (SDP), gradient-based techniques or seesaw strategy, our approach not only demonstrates superior computational efficiency but also maintains applicability across various scenarios within a streamlined workflow. We substantiate the efficacy of our method by applying it to several key quantum resources, including entanglement, coherence, and magic states. Moreover, our methodology can be readily extended to other convex roof quantities beyond the domain of resource theories, suggesting broad applicability in the realm of quantum information theory. | [
{
"version": "v1",
"created": "Fri, 28 Jun 2024 06:45:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zhu",
"Xuanran",
""
],
[
"Zhang",
"Chao",
""
],
[
"An",
"Zheng",
""
],
[
"Zeng",
"Bei",
""
]
] |
2407.00397 | Weihan Li | Weihan Li, Yule Wang, Chengrui Li, Anqi Wu | Learning Time-Varying Multi-Region Brain Communications via Scalable Markovian Gaussian Processes | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding and constructing brain communications that capture dynamic communications across multiple regions is fundamental to modern system neuroscience, yet current methods struggle to find time-varying region-level communications or scale to large neural datasets with long recording durations. We present a novel framework using Markovian Gaussian Processes to learn brain communications with time-varying temporal delays from multi-region neural recordings, named Adaptive Delay Model (ADM). Our method combines Gaussian Processes with State Space Models and employs parallel scan inference algorithms, enabling efficient scaling to large datasets while identifying concurrent communication patterns that evolve over time. This time-varying approach captures how brain region interactions shift dynamically during cognitive processes. Validated on synthetic and multi-region neural recordings datasets, our approach discovers both the directionality and temporal dynamics of neural communication. This work advances our understanding of distributed neural computation and provides a scalable tool for analyzing dynamic brain networks. | [
{
"version": "v1",
"created": "Sat, 29 Jun 2024 10:50:23 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Feb 2025 16:40:43 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Feb 2025 19:33:03 GMT"
},
{
"version": "v4",
"created": "Fri, 06 Jun 2025 11:51:23 GMT"
},
{
"version": "v5",
"created": "Wed, 11 Jun 2025 12:36:31 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Li",
"Weihan",
""
],
[
"Wang",
"Yule",
""
],
[
"Li",
"Chengrui",
""
],
[
"Wu",
"Anqi",
""
]
] |
2407.01067 | Huiguang He | Changde Du, Kaicheng Fu, Bincheng Wen, Yi Sun, Jie Peng, Wei Wei, Ying Gao, Shengpei Wang, Chuncheng Zhang, Jinpeng Li, Shuang Qiu, Le Chang, and Huiguang He | Human-like object concept representations emerge naturally in multimodal large language models | Published on Nature Machine Intelligence | Nature Machine Intelligence, 2025 | 10.1038/s42256-025-01049-z | null | cs.AI cs.CL cs.CV cs.HC cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Understanding how humans conceptualize and categorize natural objects offers critical insights into perception and cognition. With the advent of Large Language Models (LLMs), a key question arises: can these models develop human-like object representations from linguistic and multimodal data? In this study, we combined behavioral and neuroimaging analyses to explore the relationship between object concept representations in LLMs and human cognition. We collected 4.7 million triplet judgments from LLMs and Multimodal LLMs (MLLMs) to derive low-dimensional embeddings that capture the similarity structure of 1,854 natural objects. The resulting 66-dimensional embeddings were stable, predictive, and exhibited semantic clustering similar to human mental representations. Remarkably, the dimensions underlying these embeddings were interpretable, suggesting that LLMs and MLLMs develop human-like conceptual representations of objects. Further analysis showed strong alignment between model embeddings and neural activity patterns in brain regions such as EBA, PPA, RSC, and FFA. This provides compelling evidence that the object representations in LLMs, while not identical to human ones, share fundamental similarities that reflect key aspects of human conceptual knowledge. Our findings advance the understanding of machine intelligence and inform the development of more human-like artificial cognitive systems. | [
{
"version": "v1",
"created": "Mon, 01 Jul 2024 08:17:19 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 06:55:41 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 10:15:37 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Du",
"Changde",
""
],
[
"Fu",
"Kaicheng",
""
],
[
"Wen",
"Bincheng",
""
],
[
"Sun",
"Yi",
""
],
[
"Peng",
"Jie",
""
],
[
"Wei",
"Wei",
""
],
[
"Gao",
"Ying",
""
],
[
"Wang",
"Shengpei",
""
],
[
"Zhang",
"Chuncheng",
""
],
[
"Li",
"Jinpeng",
""
],
[
"Qiu",
"Shuang",
""
],
[
"Chang",
"Le",
""
],
[
"He",
"Huiguang",
""
]
] |
2407.01250 | Yang Pan | Yang Pan, Clemens Hutter, Helmut B\"olcskei | Metric-Entropy Limits on the Approximation of Nonlinear Dynamical Systems | null | null | null | null | cs.LG cs.IT math.DS math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper is concerned with fundamental limits on the approximation of nonlinear dynamical systems. Specifically, we show that recurrent neural networks (RNNs) can approximate nonlinear systems -- that satisfy a Lipschitz property and forget past inputs fast enough -- in metric-entropy-optimal manner. As the sets of sequence-to-sequence mappings realized by the dynamical systems we consider are significantly more massive than function classes generally analyzed in approximation theory, a refined metric-entropy characterization is needed, namely in terms of order, type, and generalized dimension. We compute these quantities for the classes of exponentially- and polynomially Lipschitz fading-memory systems and show that RNNs can achieve them. | [
{
"version": "v1",
"created": "Mon, 01 Jul 2024 12:57:03 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 10:51:53 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Pan",
"Yang",
""
],
[
"Hutter",
"Clemens",
""
],
[
"Bölcskei",
"Helmut",
""
]
] |
2407.04934 | Daniel Revier | Daniel Revier, Molly Carton, Jeffrey I. Lipton | Design and Reprogrammability of Zero Modes in 2D Materials from a Single Element | 14 pages, 4 figures | null | null | null | physics.app-ph physics.class-ph | http://creativecommons.org/licenses/by/4.0/ | Mechanical extremal materials, a class of metamaterials that exist at the bounds of elastic theory, possess the extraordinary capability to engineer any desired elastic behavior by harnessing mechanical zero modes -- deformation modes that demand minimal or, ideally, no elastic energy. However, the potential for arbitrary construction and reprogramming of metamaterials remains largely unrealized, primarily due to significant challenges in qualitatively transforming zero modes within the confines of existing metamaterial design frameworks. This work presents a method for explicitly defining and in situ reprogramming zero modes of two-dimensional extremal materials by employing straight-line mechanisms (SLMs) and planar symmetry, which prescribe and coordinate the zero modes, respectively. We validate the concept experimentally on square-symmetric lattices and corroborate its generality for hexagonal lattices through finite-element analysis, together spanning the full theoretical gamut of extremal behaviors. The method is used to design, test, and reprogram centimeter-scale isotropic, orthotropic, and chiral extremal materials by reorienting the SLMs in place, enabling these materials to smoothly and reversibly interpolate between extremal modalities (e.g., unimode to bimode), material properties (e.g., negative to positive Poisson's ratios), and selectively enable chirality without changing the metamaterial's global structure. This methodology provides a straightforward and explicit strategy for the design and tuning of all varieties of two-dimensional extremal materials, enabling dynamic mechanical metamaterial construction to completely cover the gamut of elastic properties. | [
{
"version": "v1",
"created": "Sat, 06 Jul 2024 02:51:25 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 19:07:37 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Revier",
"Daniel",
""
],
[
"Carton",
"Molly",
""
],
[
"Lipton",
"Jeffrey I.",
""
]
] |
2407.05277 | Tamir Bendory | Amnon Balanov, Wasim Huleihel, and Tamir Bendory | Einstein from Noise: Statistical Analysis | null | null | null | null | eess.SP math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | ``Einstein from noise" (EfN) is a prominent example of the model bias phenomenon: systematic errors in the statistical model that lead to spurious but consistent estimates. In the EfN experiment, one falsely believes that a set of observations contains noisy, shifted copies of a template signal (e.g., an Einstein image), whereas in reality, it contains only pure noise observations. To estimate the signal, the observations are first aligned with the template using cross-correlation, and then averaged. Although the observations contain nothing but noise, it was recognized early on that this process produces a signal that resembles the template signal! This pitfall was at the heart of a central scientific controversy about validation techniques in structural biology.
This paper provides a comprehensive statistical analysis of the EfN phenomenon above. We show that the Fourier phases of the EfN estimator (namely, the average of the aligned noise observations) converge to the Fourier phases of the template signal, explaining the observed structural similarity. Additionally, we prove that the convergence rate is inversely proportional to the number of noise observations and, in the high-dimensional regime, to the Fourier magnitudes of the template signal. Moreover, in the high-dimensional regime, the Fourier magnitudes converge to a scaled version of the template signal's Fourier magnitudes. This work not only deepens the theoretical understanding of the EfN phenomenon but also highlights potential pitfalls in template matching techniques and emphasizes the need for careful interpretation of noisy observations across disciplines in engineering, statistics, physics, and biology. | [
{
"version": "v1",
"created": "Sun, 07 Jul 2024 06:21:28 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 09:17:58 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Balanov",
"Amnon",
""
],
[
"Huleihel",
"Wasim",
""
],
[
"Bendory",
"Tamir",
""
]
] |
2407.06100 | Syed Zahid Husain | Syed Zahid Husain, Leo Separovic, Jean-Fran\c{c}ois Caron, Rabah Aider, Mark Buehner, St\'ephane Chamberland, Ervig Lapalme, Ron McTaggart-Cowan, Christopher Subich, Paul A. Vaillancourt, Jing Yang, Ayrton Zadra | Leveraging data-driven weather models for improving numerical weather prediction skill through large-scale spectral nudging | null | null | 10.1175/WAF-D-24-0139.1 | null | physics.ao-ph cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Operational meteorological forecasting has long relied on physics-based numerical weather prediction (NWP) models. Recently, this landscape has faced disruption by the advent of data-driven artificial intelligence (AI)-based weather models, which offer tremendous computational performance and competitive forecasting accuracy. However, data-driven models for medium-range forecasting generally suffer from major limitations, including low effective resolution and a narrow range of predicted variables. This study illustrates the relative strengths and weaknesses of these competing paradigms using the physics-based GEM (Global Environmental Multiscale) and the AI-based GraphCast models. Analyses of their respective global predictions in physical and spectral space reveal that GraphCast-predicted large scales outperform GEM, particularly for longer lead times, even though fine scales predicted by GraphCast suffer from excessive smoothing. Building on this insight, a hybrid NWP-AI system is proposed, wherein temperature and horizontal wind components predicted by GEM are spectrally nudged toward GraphCast predictions at large scales, while GEM itself freely generates the fine-scale details critical for local predictability and weather extremes. This hybrid approach is capable of leveraging the strengths of GraphCast to enhance the prediction skill of the GEM model while generating a full suite of physically consistent forecast fields with a full power spectrum. Additionally, trajectories of tropical cyclones are predicted with enhanced accuracy without significant changes in intensity. Work is in progress for operationalization of this hybrid system at the Canadian Meteorological Centre. | [
{
"version": "v1",
"created": "Mon, 08 Jul 2024 16:39:25 GMT"
},
{
"version": "v2",
"created": "Wed, 24 Jul 2024 21:23:38 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 15:48:16 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Husain",
"Syed Zahid",
""
],
[
"Separovic",
"Leo",
""
],
[
"Caron",
"Jean-François",
""
],
[
"Aider",
"Rabah",
""
],
[
"Buehner",
"Mark",
""
],
[
"Chamberland",
"Stéphane",
""
],
[
"Lapalme",
"Ervig",
""
],
[
"McTaggart-Cowan",
"Ron",
""
],
[
"Subich",
"Christopher",
""
],
[
"Vaillancourt",
"Paul A.",
""
],
[
"Yang",
"Jing",
""
],
[
"Zadra",
"Ayrton",
""
]
] |
2407.07060 | Christian Pluchar | Christian M. Pluchar, Wenhua He, Jack Manley, Nicolas Deshler, Saikat Guha, and Dalziel J. Wilson | Imaging-based Quantum Optomechanics | 10 pages, 5 figures | null | 10.1103/64xv-3fyx | null | quant-ph physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In active imaging protocols, information about an object is encoded into the spatial mode of a scattered photon. Recently the quantum limits of active imaging have been explored with levitated nanoparticles, which experience a multimode radiation pressure backaction (the photon recoil force) due to radiative scattering of the probe field. Here we extend the analysis of multimode backaction to compliant surfaces, accessing a broad class of mechanical resonators and fruitful analogies to quantum imaging. As an example, we consider imaging of the flexural modes of a membrane by sorting the spatial modes of a laser reflected from its surface. We show that backaction in this setting can be understood to arise from spatiotemporal photon shot noise, an effect that cannot be observed in single-mode optomechanics. We also derive the imprecision-backaction product in the limit of purely spatial (intermodal) coupling, revealing it to be equivalent to the standard quantum limit for single-mode optomechanical coupling. Finally, we show that optomechanical correlations due to spatiotemporal backaction can give rise to two-mode entangled light, providing a mechanism for entangling desired pairs of spatial modes. In conjunction with high-Q nanomechanics, our findings point to new opportunities at the interface of quantum imaging and optomechanics, including sensors and networks enhanced by spatial mode entanglement. | [
{
"version": "v1",
"created": "Tue, 09 Jul 2024 17:31:50 GMT"
},
{
"version": "v2",
"created": "Fri, 16 May 2025 18:02:53 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 01:39:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Pluchar",
"Christian M.",
""
],
[
"He",
"Wenhua",
""
],
[
"Manley",
"Jack",
""
],
[
"Deshler",
"Nicolas",
""
],
[
"Guha",
"Saikat",
""
],
[
"Wilson",
"Dalziel J.",
""
]
] |
2407.08792 | Dzung Pham | Dzung Pham, Jade Sheffey, Chau Minh Pham, Amir Houmansadr | ProxyGPT: Enabling User Anonymity in LLM Chatbots via (Un)Trustworthy Volunteer Proxies | null | null | null | null | cs.CR | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Popular large language model (LLM) chatbots such as ChatGPT and Claude require users to create an account with an email or a phone number before allowing full access to their services. This practice ties users' personally identifiable information (PII) to their sensitive conversational data, thus posing significant privacy risks. Unfortunately, existing private LLM solutions based on cryptography or trusted execution environments (TEEs) remain unpopular due to their prohibitive computational expense and platform restrictions. To enable practical user anonymity in LLM chatbots, we propose ProxyGPT, a privacy-enhancing system that leverages browser interaction proxies to submit user queries on their behalf. Unlike traditional proxy systems, ProxyGPT operates at the "user" layer by proxying user interactions with the browser in identity-required environments, thus easily supporting a wide range of chatbot services. We prevent malicious proxies by performing regular integrity audits using modern web proof protocols for TLS data provenance. We further utilize state-of-the-art LLM prompt guards on the proxy's side to mitigate unwanted user requests. Additionally, we incorporate a give-and-take economy based on Chaum's blind-signature e-cash to incentivize ProxyGPT users to proxy for others. Our system evaluation and user study demonstrate the practicality of our approach, as each chat request only takes a few additional seconds on average to fully complete. To the best of our knowledge, ProxyGPT is the first comprehensive proxy-based solution for privacy-preserving AI chatbots. | [
{
"version": "v1",
"created": "Thu, 11 Jul 2024 18:08:04 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 05:11:41 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Pham",
"Dzung",
""
],
[
"Sheffey",
"Jade",
""
],
[
"Pham",
"Chau Minh",
""
],
[
"Houmansadr",
"Amir",
""
]
] |
2407.09275 | Harry Petyt | Zachary Munro and Harry Petyt | Coarse obstructions to cocompact cubulation | 26 pages. v3: Added Theorem F and Remark 3.3. v2: small corrections | null | null | null | math.GR math.MG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We provide geometric methods to give bounds on the large-scale dimension of CAT(0) cube complexes quasiisometric to a given group $G$. In situations where these bounds conflict we obtain obstructions to $G$ being cocompactly cubulated. More strongly, the obstructions prevent $G$ from being a coarse median space.
As applications, we show that many free-by-cyclic groups cannot be cocompactly cubulated, even virtually, and prove that any tubular group with a coarse median is virtually compact special. We also exhibit a group that is CAT(0), $C(6)$, and virtually special, yet is not quasiisometric to any CAT(0) cube complex. This is the first example of a $C(6)$ group that cannot be cocompactly cubulated, resolving a question of Jankiewicz and partially answering a question of Wise. | [
{
"version": "v1",
"created": "Fri, 12 Jul 2024 14:03:18 GMT"
},
{
"version": "v2",
"created": "Tue, 20 Aug 2024 15:43:05 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 14:00:51 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Munro",
"Zachary",
""
],
[
"Petyt",
"Harry",
""
]
] |
2407.10454 | Amin Rakhsha | Jongmin Lee, Amin Rakhsha, Ernest K. Ryu, Amir-massoud Farahmand | Deflated Dynamics Value Iteration | null | Transactions on Machine Learning Research 2025 | null | null | cs.LG math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Value Iteration (VI) algorithm is an iterative procedure to compute the value function of a Markov decision process, and is the basis of many reinforcement learning (RL) algorithms as well. As the error convergence rate of VI as a function of iteration $k$ is $O(\gamma^k)$, it is slow when the discount factor $\gamma$ is close to $1$. To accelerate the computation of the value function, we propose Deflated Dynamics Value Iteration (DDVI). DDVI uses matrix splitting and matrix deflation techniques to effectively remove (deflate) the top $s$ dominant eigen-structure of the transition matrix $\mathcal{P}^{\pi}$. We prove that this leads to a $\tilde{O}(\gamma^k |\lambda_{s+1}|^k)$ convergence rate, where $\lambda_{s+1}$is $(s+1)$-th largest eigenvalue of the dynamics matrix. We then extend DDVI to the RL setting and present Deflated Dynamics Temporal Difference (DDTD) algorithm. We empirically show the effectiveness of the proposed algorithms. | [
{
"version": "v1",
"created": "Mon, 15 Jul 2024 06:07:05 GMT"
},
{
"version": "v2",
"created": "Sat, 31 May 2025 14:46:15 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lee",
"Jongmin",
""
],
[
"Rakhsha",
"Amin",
""
],
[
"Ryu",
"Ernest K.",
""
],
[
"Farahmand",
"Amir-massoud",
""
]
] |
2407.12736 | Mohammad Erfan Sadeghi | Mohammad Erfan Sadeghi, Arash Fayyazi, Suhas Somashekar, Armin Abdollahi, Massoud Pedram | CHOSEN: Compilation to Hardware Optimization Stack for Efficient Vision Transformer Inference | null | null | null | null | cs.CV cs.AI cs.AR | http://creativecommons.org/licenses/by/4.0/ | Vision Transformers (ViTs) represent a groundbreaking shift in machine learning approaches to computer vision. Unlike traditional approaches, ViTs employ the self-attention mechanism, which has been widely used in natural language processing, to analyze image patches. Despite their advantages in modeling visual tasks, deploying ViTs on hardware platforms, notably Field-Programmable Gate Arrays (FPGAs), introduces considerable challenges. These challenges stem primarily from the non-linear calculations and high computational and memory demands of ViTs. This paper introduces CHOSEN, a software-hardware co-design framework to address these challenges and offer an automated framework for ViT deployment on the FPGAs in order to maximize performance. Our framework is built upon three fundamental contributions: multi-kernel design to maximize the bandwidth, mainly targeting benefits of multi DDR memory banks, approximate non-linear functions that exhibit minimal accuracy degradation, and efficient use of available logic blocks on the FPGA, and efficient compiler to maximize the performance and memory-efficiency of the computing kernels by presenting a novel algorithm for design space exploration to find optimal hardware configuration that achieves optimal throughput and latency. Compared to the state-of-the-art ViT accelerators, CHOSEN achieves a 1.5x and 1.42x improvement in the throughput on the DeiT-S and DeiT-B models. | [
{
"version": "v1",
"created": "Wed, 17 Jul 2024 16:56:06 GMT"
},
{
"version": "v2",
"created": "Mon, 22 Jul 2024 05:19:15 GMT"
},
{
"version": "v3",
"created": "Thu, 25 Jul 2024 00:00:18 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 18:33:53 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Sadeghi",
"Mohammad Erfan",
""
],
[
"Fayyazi",
"Arash",
""
],
[
"Somashekar",
"Suhas",
""
],
[
"Abdollahi",
"Armin",
""
],
[
"Pedram",
"Massoud",
""
]
] |
2407.13329 | Lorenzo Paolini | Lorenzo Paolini, Sahar Vahdati, Angelo Di Iorio, Robert Wardenga, Ivan Heibi, Silvio Peroni | CiteFusion: An Ensemble Framework for Citation Intent Classification Harnessing Dual-Model Binary Couples and SHAP Analyses | Submitted to Scientometrics Journal | null | 10.5281/zenodo.15011985 | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding the motivations underlying scholarly citations is essential to evaluate research impact and pro-mote transparent scholarly communication. This study introduces CiteFusion, an ensemble framework designed to address the multi-class Citation Intent Classification task on two benchmark datasets: SciCite and ACL-ARC. The framework employs a one-vs-all decomposition of the multi-class task into class-specific binary sub-tasks, leveraging complementary pairs of SciBERT and XLNet models, independently tuned, for each citation intent. The outputs of these base models are aggregated through a feedforward neural network meta-classifier to reconstruct the original classification task. To enhance interpretability, SHAP (SHapley Additive exPlanations) is employed to analyze token-level contributions, and interactions among base models, providing transparency into the classification dynamics of CiteFusion, and insights about the kind of misclassifications of the ensem-ble. In addition, this work investigates the semantic role of structural context by incorporating section titles, as framing devices, into input sentences, assessing their positive impact on classification accuracy. CiteFusion ul-timately demonstrates robust performance in imbalanced and data-scarce scenarios: experimental results show that CiteFusion achieves state-of-the-art performance, with Macro-F1 scores of 89.60% on SciCite, and 76.24% on ACL-ARC. Furthermore, to ensure interoperability and reusability, citation intents from both datasets sche-mas are mapped to Citation Typing Ontology (CiTO) object properties, highlighting some overlaps. Finally, we describe and release a web-based application that classifies citation intents leveraging the CiteFusion models developed on SciCite. | [
{
"version": "v1",
"created": "Thu, 18 Jul 2024 09:29:33 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Mar 2025 11:59:18 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:16:14 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Paolini",
"Lorenzo",
""
],
[
"Vahdati",
"Sahar",
""
],
[
"Di Iorio",
"Angelo",
""
],
[
"Wardenga",
"Robert",
""
],
[
"Heibi",
"Ivan",
""
],
[
"Peroni",
"Silvio",
""
]
] |
2407.13625 | Aras Selvi | Aras Selvi, Eleonora Kreacic, Mohsen Ghassemi, Vamsi Potluru, Tucker Balch, Manuela Veloso | Distributionally and Adversarially Robust Logistic Regression via Intersecting Wasserstein Balls | 9 main pages + 25 pages of appendices | null | null | null | math.OC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adversarially robust optimization (ARO) has emerged as the *de facto* standard for training models that hedge against adversarial attacks in the test stage. While these models are robust against adversarial attacks, they tend to suffer severely from overfitting. To address this issue, some successful methods replace the empirical distribution in the training stage with alternatives including *(i)* a worst-case distribution residing in an ambiguity set, resulting in a distributionally robust (DR) counterpart of ARO; *(ii)* a mixture of the empirical distribution with a distribution induced by an auxiliary (*e.g.*, synthetic, external, out-of-domain) dataset. Inspired by the former, we study the Wasserstein DR counterpart of ARO for logistic regression and show it admits a tractable convex optimization reformulation. Adopting the latter setting, we revise the DR approach by intersecting its ambiguity set with another ambiguity set built using the auxiliary dataset, which offers a significant improvement whenever the Wasserstein distance between the data generating and auxiliary distributions can be estimated. We study the underlying optimization problem, develop efficient solution algorithms, and demonstrate that the proposed method outperforms benchmark approaches on standard datasets. | [
{
"version": "v1",
"created": "Thu, 18 Jul 2024 15:59:37 GMT"
},
{
"version": "v2",
"created": "Fri, 18 Oct 2024 11:43:28 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 14:59:55 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 12:08:46 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Selvi",
"Aras",
""
],
[
"Kreacic",
"Eleonora",
""
],
[
"Ghassemi",
"Mohsen",
""
],
[
"Potluru",
"Vamsi",
""
],
[
"Balch",
"Tucker",
""
],
[
"Veloso",
"Manuela",
""
]
] |
2407.13817 | Benoit Laurent | Gonzalo Alonso-\'Alvarez, James M. Cline, Benoit Laurent, Ushak Rahaman | Dark photon distortions of NO$\nu$A and T2K neutrino oscillations | 7 pages, 3 figure; v3: Added figures 1 and 3 | null | null | CERN-TH-2024-114 | hep-ph astro-ph.CO hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dark photons coupling to $L_\mu-L_\tau$ lepton number difference are a highly studied light dark matter candidate, with potential to be discovered through their impact on terrestrial neutrino oscillation experiments. We re-examine this in the light of claimed tensions between the NO$\nu$A and T2K long baseline experiments, also taking into account data from the MINOS experiment. We obtain leading limits on the $L_\mu-L_\tau$ gauge coupling $g'$ versus dark photon mass $m_{A'}$, and find no statistically significant alleviation of the tension from inclusion of the new physics effect. | [
{
"version": "v1",
"created": "Thu, 18 Jul 2024 18:00:00 GMT"
},
{
"version": "v2",
"created": "Tue, 30 Jul 2024 18:55:07 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 19:48:48 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Alonso-Álvarez",
"Gonzalo",
""
],
[
"Cline",
"James M.",
""
],
[
"Laurent",
"Benoit",
""
],
[
"Rahaman",
"Ushak",
""
]
] |
2407.14782 | Arian Vezvaee | Arian Vezvaee, Vinay Tripathi, Daria Kowsari, Eli Levenson-Falk, Daniel A. Lidar | Virtual Z gates and symmetric gate compilation | null | null | 10.1103/PRXQuantum.6.020348 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | The virtual Z gate has been established as an important tool for performing quantum gates on various platforms, including but not limited to superconducting systems. Many such platforms offer a limited set of calibrated gates and compile all other gates using combinations of X-type and virtual Z gates. Here, we show that the method of compilation has important consequences in an open quantum system setting. Specifically, we experimentally demonstrate that it is crucial to choose a compilation that is symmetric with respect to virtual Z rotations. An important example is dynamical decoupling (DD) sequences, where improper gate decomposition can result in unintended effects such as the implementation of the wrong sequence. Our findings indicate that in many cases the performance of DD is adversely affected by the incorrect use of virtual Z gates, compounding other coherent pulse errors. This holds even for DD sequences designed to be robust against systematic control errors. In addition, we identify another source of coherent errors: interference between consecutive pulses that follow each other too closely. This work provides insights into improving general quantum gate performance and optimizing DD sequences in particular. | [
{
"version": "v1",
"created": "Sat, 20 Jul 2024 07:07:08 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Mar 2025 23:22:47 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Vezvaee",
"Arian",
""
],
[
"Tripathi",
"Vinay",
""
],
[
"Kowsari",
"Daria",
""
],
[
"Levenson-Falk",
"Eli",
""
],
[
"Lidar",
"Daniel A.",
""
]
] |
2407.16239 | Ahmet Zahid Balc{\i}o\u{g}lu | Ahmet Zahid Balc{\i}o\u{g}lu, Newton Mwai, Emil Carlsson, Fredrik D. Johansson | Identifiable Latent Bandits: Leveraging observational data for personalized decision-making | 30 pages, 16 figures | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For many decision-making tasks, such as precision medicine, historical data alone are insufficient to determine the right choice for a new problem instance or patient. Online algorithms like multi-armed bandits can find optimal personalized decisions but are notoriously sample-hungry. In practice, training a bandit for a new individual from scratch is often infeasible, as the number of trials required is larger than the practical number of decision points. Latent bandits offer rapid exploration and personalization beyond what context variables can reveal, provided that a latent variable model can be learned consistently. In this work, we propose an identifiable latent bandit framework that leads to optimal decision-making with a shorter exploration time than classical bandits by learning from historical records of decisions and outcomes. Our method is based on nonlinear independent component analysis that provably identifies representations from observational data sufficient to infer the optimal action in new bandit instances. We verify this strategy in simulated and semi-synthetic environments, showing substantial improvement over online and offline learning baselines when identifying conditions are satisfied. | [
{
"version": "v1",
"created": "Tue, 23 Jul 2024 07:26:38 GMT"
},
{
"version": "v2",
"created": "Mon, 29 Jul 2024 14:04:20 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 08:30:20 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 09:30:45 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Balcıoğlu",
"Ahmet Zahid",
""
],
[
"Mwai",
"Newton",
""
],
[
"Carlsson",
"Emil",
""
],
[
"Johansson",
"Fredrik D.",
""
]
] |
2407.17152 | Yuyan Chen | Yuyan Chen, Songzhou Yan, Zhihong Zhu, Zhixu Li, Yanghua Xiao | XMeCap: Meme Caption Generation with Sub-Image Adaptability | Accepted to ACM Multimedia 2024 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Humor, deeply rooted in societal meanings and cultural details, poses a unique challenge for machines. While advances have been made in natural language processing, real-world humor often thrives in a multi-modal context, encapsulated distinctively by memes. This paper poses a particular emphasis on the impact of multi-images on meme captioning. After that, we introduce the \textsc{XMeCap} framework, a novel approach that adopts supervised fine-tuning and reinforcement learning based on an innovative reward model, which factors in both global and local similarities between visuals and text. Our results, benchmarked against contemporary models, manifest a marked improvement in caption generation for both single-image and multi-image memes, as well as different meme categories. \textsc{XMeCap} achieves an average evaluation score of 75.85 for single-image memes and 66.32 for multi-image memes, outperforming the best baseline by 6.75\% and 8.56\%, respectively. This research not only establishes a new frontier in meme-related studies but also underscores the potential of machines in understanding and generating humor in a multi-modal setting. | [
{
"version": "v1",
"created": "Wed, 24 Jul 2024 10:51:46 GMT"
},
{
"version": "v2",
"created": "Wed, 31 Jul 2024 12:56:22 GMT"
},
{
"version": "v3",
"created": "Fri, 20 Sep 2024 09:39:07 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 13:23:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Yuyan",
""
],
[
"Yan",
"Songzhou",
""
],
[
"Zhu",
"Zhihong",
""
],
[
"Li",
"Zhixu",
""
],
[
"Xiao",
"Yanghua",
""
]
] |
2407.17366 | Tom H. Koornwinder | Tom H. Koornwinder and Marta Mazzocco | Automorphisms of the DAHA of type $\check{C_1}C_1$ and non-symmetric Askey-Wilson functions | v3: 39 pages, minor corrections, to appear in Indag. Math | null | 10.1016/j.indag.2025.05.005 | null | math.CA math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we consider the automorphisms of the double affine Hecke algebra (DAHA) of type $\check{C_1}C_1$ which have a relatively simple action on the generators and on the parameters, notably a symmetry $t_4$ which sends the Askey-Wilson parameters $(a,b,c,d)$ to $(a,b,qd^{-1},qc^{-1})$. We study how these symmetries act on the basic representation and on the symmetric and non-symmetric Askey-Wilson (AW) polynomials and functions. Interestingly $t_4$ maps AW polynomials to functions. We take the rank one case of Stokman's Cherednik kernel for $BC_n$ as the definition of the non-symmetric Askey--Wilson function. From it we derive an expression as a sum of a symmetric and an anti-symmetric term. | [
{
"version": "v1",
"created": "Wed, 24 Jul 2024 15:43:30 GMT"
},
{
"version": "v2",
"created": "Tue, 19 Nov 2024 16:12:11 GMT"
},
{
"version": "v3",
"created": "Thu, 29 May 2025 08:58:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Koornwinder",
"Tom H.",
""
],
[
"Mazzocco",
"Marta",
""
]
] |
2407.17718 | Shi-Shun Chen | Shi-Shun Chen, Xiao-Yang Li | Comparison of global sensitivity analysis methods for a fire spread model with a segmented characteristic | null | Mathematics and Computers in Simulation, 2024: (229)304-318 | 10.1016/j.matcom.2024.10.012 | null | stat.AP | http://creativecommons.org/licenses/by/4.0/ | Global sensitivity analysis (GSA) can provide rich information for controlling output uncertainty. In practical applications, segmented models are commonly used to describe an abrupt model change. For segmented models, the complicated uncertainty propagation during the transition region may lead to different importance rankings of different GSA methods. If an unsuitable GSA method is applied, misleading results will be obtained, resulting in suboptimal or even wrong decisions. In this paper, four GSA indices, i.e., Sobol index, mutual information, delta index and PAWN index, are applied for a segmented fire spread model (Dry Eucalypt). The results show that four GSA indices give different importance rankings during the transition region since segmented characteristics affect different GSA indices in different ways. We suggest that analysts should rely on the results of different GSA indices according to their practical purpose, especially when making decisions for segmented models during the transition region. All of our source codes are publicly available at https://github.com/dirge1/GSA_segmented. | [
{
"version": "v1",
"created": "Thu, 25 Jul 2024 02:40:38 GMT"
},
{
"version": "v2",
"created": "Mon, 03 Mar 2025 11:53:25 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 18:18:06 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Shi-Shun",
""
],
[
"Li",
"Xiao-Yang",
""
]
] |
2407.17719 | Shi-Shun Chen | Shi-Shun Chen, Xiao-Yang Li | A new moment-independent uncertainty importance measure based on cumulative residual entropy for developing uncertainty reduction strategies | null | Mathematics and Computers in Simulation 2025 | 10.1016/j.matcom.2025.06.004 | null | stat.AP | http://creativecommons.org/licenses/by/4.0/ | Uncertainty reduction is vital for improving system reliability and reducing risks. To identify the best target for uncertainty reduction, uncertainty importance measure is commonly used to prioritize the significance of input variable uncertainties. Then, designers will take steps to reduce the uncertainties of variables with high importance. However, for variables with minimal uncertainty, the cost of controlling their uncertainties can be unacceptable. Therefore, uncertainty magnitude should also be considered in developing uncertainty reduction strategies. Although variance-based methods have been developed for this purpose, they are dependent on statistical moments and have limitations when dealing with highly-skewed distributions that are commonly encountered in practical applications. Motivated by this problem, we propose a new uncertainty importance measure based on cumulative residual entropy. The proposed measure is moment-independent based on the cumulative distribution function, which can handle the highly-skewed distributions properly. Numerical implementations for estimating the proposed measure are devised and verified. A real-world engineering case considering highly-skewed distributions is introduced to show the procedure of developing uncertainty reduction strategies considering uncertainty magnitude and corresponding cost. The results demonstrate that the proposed measure can present a different uncertainty reduction recommendation compared to the variance-based approach because of its moment-independent characteristic. | [
{
"version": "v1",
"created": "Thu, 25 Jul 2024 02:43:04 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 18:27:40 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Shi-Shun",
""
],
[
"Li",
"Xiao-Yang",
""
]
] |
2407.18337 | Alexandre Bousse | Youness Mellak, Alexandre Bousse, Thibaut Merlin, Debora Giovagnoli, Dimitris Visvikis | Direct3{\gamma}: A Pipeline for Direct Three-gamma PET Image Reconstruction | 11 pages, 11 figures, 2 tables | null | 10.1109/TRPMS.2025.3577810 | null | physics.med-ph | http://creativecommons.org/licenses/by/4.0/ | This paper presents a novel image reconstruction pipeline for three-gamma (3-{\gamma}) positron emission tomography (PET) aimed at improving spatial resolution and reducing noise in nuclear medicine. The proposed Direct3{\gamma} pipeline addresses the inherent challenges in 3-{\gamma} PET systems, such as detector imperfections and uncertainty in photon interaction points. A key feature of the pipeline is its ability to determine the order of interactions through a model trained on Monte Carlo (MC) simulations using the Geant4 Application for Tomography Emission (GATE) toolkit, thus providing the necessary information to construct Compton cones which intersect with the line of response (LOR) to provide an estimate of the emission point. The pipeline processes 3-{\gamma} PET raw data, reconstructs histoimages by propagating energy and spatial uncertainties along the LOR, and applies a 3-D convolutional neural network (CNN) to refine these intermediate images into high-quality reconstructions. To further enhance image quality, the pipeline leverages both supervised learning and adversarial losses, the latter preserving fine structural details. Experimental results show that Direct3{\gamma} consistently outperforms conventional 200-ps time-of-flight (TOF) PET in terms of SSIM and PSNR. | [
{
"version": "v1",
"created": "Thu, 25 Jul 2024 18:51:47 GMT"
},
{
"version": "v2",
"created": "Wed, 09 Oct 2024 07:01:53 GMT"
},
{
"version": "v3",
"created": "Tue, 11 Mar 2025 17:21:13 GMT"
},
{
"version": "v4",
"created": "Wed, 09 Apr 2025 11:57:08 GMT"
},
{
"version": "v5",
"created": "Thu, 05 Jun 2025 14:37:22 GMT"
},
{
"version": "v6",
"created": "Wed, 11 Jun 2025 08:18:19 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Mellak",
"Youness",
""
],
[
"Bousse",
"Alexandre",
""
],
[
"Merlin",
"Thibaut",
""
],
[
"Giovagnoli",
"Debora",
""
],
[
"Visvikis",
"Dimitris",
""
]
] |
2407.18986 | Shi-Shun Chen | Shi-Shun Chen, Yu-Tong Jiang, Wen-Bin Chen, Xiao-Yang Li | TERIME: An improved RIME algorithm with enhanced exploration and exploitation for robust parameter extraction of photovoltaic models | null | Journal of Bionic Engineering (22) 1535-1556 2025 | 10.1007/s42235-025-00679-8 | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Parameter extraction of photovoltaic (PV) models is crucial for the planning, optimization, and control of PV systems. Although some methods using meta-heuristic algorithms have been proposed to determine these parameters, the robustness of solutions obtained by these methods faces great challenges when the complexity of the PV model increases. The unstable results will affect the reliable operation and maintenance strategies of PV systems. In response to this challenge, an improved rime optimization algorithm with enhanced exploration and exploitation, termed TERIME, is proposed for robust and accurate parameter identification for various PV models. Specifically, the differential evolution mutation operator is integrated in the exploration phase to enhance the population diversity. Meanwhile, a new exploitation strategy incorporating randomization and neighborhood strategies simultaneously is developed to maintain the balance of exploitation width and depth. The TERIME algorithm is applied to estimate the optimal parameters of the single diode model, double diode model, and triple diode model combined with the Lambert-W function for three PV cell and module types including RTC France, Photo Watt-PWP 201 and S75. According to the statistical analysis in 100 runs, the proposed algorithm achieves more accurate and robust parameter estimations than other techniques to various PV models in varying environmental conditions. All of our source codes are publicly available at https://github.com/dirge1/TERIME. | [
{
"version": "v1",
"created": "Thu, 25 Jul 2024 02:47:14 GMT"
},
{
"version": "v2",
"created": "Thu, 01 Aug 2024 15:59:39 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 18:26:25 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Shi-Shun",
""
],
[
"Jiang",
"Yu-Tong",
""
],
[
"Chen",
"Wen-Bin",
""
],
[
"Li",
"Xiao-Yang",
""
]
] |
2407.19917 | George Mihailescu | George Mihailescu, Steve Campbell, Karol Gietka | Uncertain Quantum Critical Metrology: From Single to Multi Parameter Sensing | 9 pages, 3 figures | Phys. Rev. A 111, 052621, 2025 | 10.1103/PhysRevA.111.052621 | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Critical quantum metrology relies on the extreme sensitivity of a system's eigenstates near the critical point of a quantum phase transition to Hamiltonian perturbations. This means that these eigenstates are extremely sensitive to all the parameters of the Hamiltonian. In realistic settings, there is always some degree of uncertainty in the control parameters used to tune the system to criticality. These uncertainties, while not the target of estimation, can significantly affect the attainable precision, effectively acting as nuisance parameters in the estimation process. Despite being a practically relevant source of noise, their impact on critical metrology has been largely overlooked. In this work we present a general framework that interpolates between single- and multiparameter estimation settings, enabling a systematic analysis of how such uncertainties influence sensitivity. We apply this framework to the paradigmatic transverse field Ising and Lipkin-Meshkov-Glick models, explicitly demonstrating how uncertainty in control parameters affects the metrological performance of critical sensors. For finite-size systems, we identify a fundamental trade-off between robustness to uncertainty and the ability to retain a quantum advantage at the critical point. Our results contribute to a deeper understanding of the practical limitations of critical quantum metrology and provide a route toward its more resilient implementation. | [
{
"version": "v1",
"created": "Mon, 29 Jul 2024 11:50:21 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:25:12 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Mihailescu",
"George",
""
],
[
"Campbell",
"Steve",
""
],
[
"Gietka",
"Karol",
""
]
] |
2407.20165 | Sunbochen Tang | Sunbochen Tang, Haoyuan Sun, and Navid Azizan | Meta-Learning for Adaptive Control with Automated Mirror Descent | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Adaptive control achieves concurrent parameter learning and stable control under uncertainties that are linearly parameterized with known nonlinear features. Nonetheless, it is often difficult to obtain such nonlinear features. To address this difficulty, recent progress has been made in integrating meta-learning with adaptive control to learn such nonlinear features from data. However, these meta-learning-based control methods rely on classical adaptation laws using gradient descent, which is confined to the Euclidean geometry. In this paper, we propose a novel method that combines meta-learning and adaptation laws based on mirror descent, a popular generalization of gradient descent, which takes advantage of the potentially non-Euclidean geometry of the parameter space. In our approach, meta-learning not only learns the nonlinear features but also searches for a suitable mirror-descent potential function that optimizes control performance. Through numerical simulations, we demonstrate the effectiveness of the proposed method in learning efficient representations and real-time tracking control performance under uncertain dynamics. | [
{
"version": "v1",
"created": "Mon, 29 Jul 2024 16:51:28 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Dec 2024 23:30:25 GMT"
},
{
"version": "v3",
"created": "Mon, 02 Jun 2025 02:01:16 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 20:16:20 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tang",
"Sunbochen",
""
],
[
"Sun",
"Haoyuan",
""
],
[
"Azizan",
"Navid",
""
]
] |
2408.00241 | Minheng Xiao | Minheng Xiao, Zhizhong Wu | Multiple Greedy Quasi-Newton Methods for Saddle Point Problems | Accepted by DOCS 2024 | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | This paper introduces the Multiple Greedy Quasi-Newton (MGSR1-SP) method, a novel approach to solving strongly-convex-strongly-concave (SCSC) saddle point problems. Our method enhances the approximation of the squared indefinite Hessian matrix inherent in these problems, significantly improving both stability and efficiency through iterative greedy updates. We provide a thorough theoretical analysis of MGSR1-SP, demonstrating its linear-quadratic convergence rate. Numerical experiments conducted on AUC maximization and adversarial debiasing problems, compared with state-of-the-art algorithms, underscore our method's enhanced convergence rate. These results affirm the potential of MGSR1-SP to improve performance across a broad spectrum of machine learning applications where efficient and accurate Hessian approximations are crucial. | [
{
"version": "v1",
"created": "Thu, 01 Aug 2024 02:40:37 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Jan 2025 02:49:43 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:52:51 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xiao",
"Minheng",
""
],
[
"Wu",
"Zhizhong",
""
]
] |
2408.00273 | Hai Shu | Yanbing Chen, Tianze Tang, Taehyo Kim, Hai Shu | UKAN-EP: Enhancing U-KAN with Efficient Attention and Pyramid Aggregation for 3D Multi-Modal MRI Brain Tumor Segmentation | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gliomas are among the most common malignant brain tumors and are characterized by considerable heterogeneity, which complicates accurate detection and segmentation. Multi-modal MRI is the clinical standard for glioma imaging, but variability across modalities and high computational complexity hinder effective automated segmentation. In this paper, we propose UKAN-EP, a novel 3D extension of the original 2D U-KAN model for multi-modal MRI brain tumor segmentation. While U-KAN integrates Kolmogorov-Arnold Network (KAN) layers into a U-Net backbone, UKAN-EP further incorporates Efficient Channel Attention (ECA) and Pyramid Feature Aggregation (PFA) modules to enhance inter-modality feature fusion and multi-scale feature representation. We also introduce a dynamic loss weighting strategy that adaptively balances the Cross-Entropy and Dice losses during training. We evaluate UKAN-EP on the 2024 BraTS-GLI dataset and compare it against strong baselines including U-Net, Attention U-Net, and Swin UNETR. Results show that UKAN-EP achieves superior segmentation performance while requiring substantially fewer computational resources. An extensive ablation study further demonstrates the effectiveness of ECA and PFA, as well as the limited utility of self-attention and spatial attention alternatives. Code is available at https://github.com/TianzeTang0504/UKAN-EP. | [
{
"version": "v1",
"created": "Thu, 01 Aug 2024 04:27:10 GMT"
},
{
"version": "v2",
"created": "Tue, 10 Jun 2025 19:14:42 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Chen",
"Yanbing",
""
],
[
"Tang",
"Tianze",
""
],
[
"Kim",
"Taehyo",
""
],
[
"Shu",
"Hai",
""
]
] |
2408.03192 | Paul-Hermann Balduf | Paul-Hermann Balduf and Davide Gaiotto | Combinatorial proof of a Non-Renormalization Theorem | 42 pages | J. High Energ. Phys. 05 (2025), 120 | 10.1007/JHEP05(2025)120 | null | math-ph hep-th math.CO math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We provide a direct combinatorial proof of a Feynman graph identity which implies a wide generalization of a formality theorem by Kontsevich. For a Feynman graph $\Gamma$, we associate to each vertex a position $x_v \in \mathbb R$ and to each edge $e$ the combination $s_e = a_e^{-\frac 12} \left( x^+_e - x^-_e \right)$, where $x^\pm_e$ are the positions of the two end vertices of $e$, and $a_e$ is a Schwinger parameter. The "topological propagator" $P_e = e^{-s_e^2}\text d s_e$ includes a part proportional to $\text d x_v$ and a part proportional to $\text d a_e$. Integrating the product of all $P_e$ over positions produces a differential form $\alpha_\Gamma$ in the variables $a_e$. We derive an explicit combinatorial formula for $\alpha_\Gamma$, and we prove that $\alpha_\Gamma \wedge \alpha_\Gamma=0$. | [
{
"version": "v1",
"created": "Tue, 06 Aug 2024 13:45:08 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Oct 2024 15:45:50 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Balduf",
"Paul-Hermann",
""
],
[
"Gaiotto",
"Davide",
""
]
] |
2408.03199 | Davide Pucci | Matteo Lapucci, Davide Pucci | Convergence Conditions for Stochastic Line Search Based Optimization of Over-parametrized Models | null | null | null | null | math.OC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we deal with algorithms to solve the finite-sum problems related to fitting over-parametrized models, that typically satisfy the interpolation condition. In particular, we focus on approaches based on stochastic line searches and employing general search directions. We define conditions on the sequence of search directions that guarantee finite termination and bounds for the backtracking procedure. Moreover, we shed light on the additional property of directions needed to prove fast (linear) convergence of the general class of algorithms when applied to PL functions in the interpolation regime. From the point of view of algorithms design, the proposed analysis identifies safeguarding conditions that could be employed in relevant algorithmic frameworks. In particular, it could be of interest to integrate stochastic line searches within momentum, conjugate gradient or adaptive preconditioning methods. | [
{
"version": "v1",
"created": "Tue, 06 Aug 2024 13:58:37 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 07:48:15 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Lapucci",
"Matteo",
""
],
[
"Pucci",
"Davide",
""
]
] |
2408.03573 | Yuheng Huang | Yuheng Huang, Jiayang Song, Qiang Hu, Felix Juefei-Xu, Lei Ma | AcTracer: Active Testing of Large Language Model via Multi-Stage Sampling | To appear in ACM Transactions on Software Engineering and Methodology (2025) | null | null | null | cs.SE cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Performance evaluation plays a crucial role in the development life cycle of large language models (LLMs). It estimates the model's capability, elucidates behavior characteristics, and facilitates the identification of potential issues and limitations, thereby guiding further improvement. Given that LLMs' diverse task-handling abilities stem from large volumes of training data, a comprehensive evaluation also necessitates abundant, well-annotated, and representative test data to assess LLM performance across various downstream tasks. However, the demand for high-quality test data often entails substantial time, computational resources, and manual efforts, sometimes causing the evaluation to be inefficient or impractical. To address these challenges, researchers propose active testing, which estimates the overall performance by selecting a subset of test data. Nevertheless, the existing active testing methods tend to be inefficient, even inapplicable, given the unique new challenges of LLMs (e.g., diverse task types, increased model complexity, and unavailability of training data). To mitigate such limitations and expedite the development cycle of LLMs, in this work, we introduce AcTracer, an active testing framework tailored for LLMs that strategically selects a small subset of test data to achieve a more accurate performance estimation for LLMs. AcTracer utilizes both internal and external information from LLMs to guide the test sampling process, reducing variance through a multi-stage pool-based active selection. Our experiment results demonstrate that AcTracer achieves state-of-the-art performance compared to existing methods across various tasks. | [
{
"version": "v1",
"created": "Wed, 07 Aug 2024 06:17:48 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 05:57:45 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Huang",
"Yuheng",
""
],
[
"Song",
"Jiayang",
""
],
[
"Hu",
"Qiang",
""
],
[
"Juefei-Xu",
"Felix",
""
],
[
"Ma",
"Lei",
""
]
] |
2408.04211 | Zhicheng Ding | Jiahao Tian and Jinman Zhao and Zhenkai Wang and Zhicheng Ding | MMREC: LLM Based Multi-Modal Recommender System | null | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | The importance of recommender systems is growing rapidly due to the exponential increase in the volume of content generated daily. This surge in content presents unique challenges for designing effective recommender systems. Key among these challenges is the need to effectively leverage the vast amounts of natural language data and images that represent user preferences. This paper presents a novel approach to enhancing recommender systems by leveraging Large Language Models (LLMs) and deep learning techniques. The proposed framework aims to improve the accuracy and relevance of recommendations by incorporating multi-modal information processing and by the use of unified latent space representation. The study explores the potential of LLMs to better understand and utilize natural language data in recommendation contexts, addressing the limitations of previous methods. The framework efficiently extracts and integrates text and image information through LLMs, unifying diverse modalities in a latent space to simplify the learning process for the ranking model. Experimental results demonstrate the enhanced discriminative power of the model when utilizing multi-modal information. This research contributes to the evolving field of recommender systems by showcasing the potential of LLMs and multi-modal data integration to create more personalized and contextually relevant recommendations. | [
{
"version": "v1",
"created": "Thu, 08 Aug 2024 04:31:29 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 04:27:29 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Tian",
"Jiahao",
""
],
[
"Zhao",
"Jinman",
""
],
[
"Wang",
"Zhenkai",
""
],
[
"Ding",
"Zhicheng",
""
]
] |
2408.04418 | Fabio Anselmi | Giuseppe D'Auria, Giovanna Morigi, Fabio Anselmi, Fabio Benatti | Decoherence Cancellation through Noise Interference | V2 Significantly re-worked version | null | null | null | quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We propose a novel, feedback-free method to cancel the effects of decoherence in the dynamics of open quantum systems subject to dephasing. The protocol makes use of the coupling with an auxiliary system when they are both subject to the same noisy dynamics, in such a way that their interaction leads to cancellation of the noise on the system itself. This requires tuning the strength of the coupling between main and auxiliary systems as well as the ability to prepare the auxiliary system in a Fock state, which solely depends on the coupling strength. We investigate the protocol's efficiency to protect NOON states against dephasing in setups such as tweezers arrays of cold atoms. We show that the protocol's efficiency is robust against fluctuations of the optimal parameters and, remarkably, that it is independent of the temporal noise features. Therefore, it can be applied to cancel both Markovian and non-Markovian noise effects, in the regime where error-correction protocols become inefficient. | [
{
"version": "v1",
"created": "Thu, 08 Aug 2024 12:45:15 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 06:58:14 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"D'Auria",
"Giuseppe",
""
],
[
"Morigi",
"Giovanna",
""
],
[
"Anselmi",
"Fabio",
""
],
[
"Benatti",
"Fabio",
""
]
] |
2408.05860 | Minheng Xiao | Minheng Xiao | Root Cause Attribution of Delivery Risks via Causal Discovery with Reinforcement Learning | null | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper presents a novel approach to root cause attribution of delivery risks within supply chains by integrating causal discovery with reinforcement learning. As supply chains become increasingly complex, traditional methods of root cause analysis struggle to capture the intricate interrelationships between various factors, often leading to spurious correlations and suboptimal decision-making. Our approach addresses these challenges by leveraging causal discovery to identify the true causal relationships between operational variables, and reinforcement learning to iteratively refine the causal graph. This method enables the accurate identification of key drivers of late deliveries, such as shipping mode and delivery status, and provides actionable insights for optimizing supply chain performance. We apply our approach to a real-world supply chain dataset, demonstrating its effectiveness in uncovering the underlying causes of delivery delays and offering strategies for mitigating these risks. The findings have significant implications for improving operational efficiency, customer satisfaction, and overall profitability within supply chains. | [
{
"version": "v1",
"created": "Sun, 11 Aug 2024 20:52:51 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Jan 2025 04:04:43 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:56:50 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xiao",
"Minheng",
""
]
] |
2408.06894 | Aidan Li | Aidan Li, Liyan Wang, Tianye Dou, Jeffrey S. Rosenthal | Exploring the generalizability of the optimal 0.234 acceptance rate in random-walk Metropolis and parallel tempering algorithms | Under review at Communications in Statistics - Simulation and Computation | null | null | null | stat.CO | http://creativecommons.org/licenses/by/4.0/ | For random-walk Metropolis (RWM) and parallel tempering (PT) algorithms, an asymptotic acceptance rate of around 0.234 is known to be optimal in certain high-dimensional limits. However, its practical relevance is uncertain due to restrictive derivation conditions. We synthesise previous theoretical advances in extending the 0.234 acceptance rate to more general settings, and demonstrate its applicability with a comprehensive empirical simulation study on examples examining how acceptance rates affect Expected Squared Jumping Distance (ESJD). Our experiments show the optimality of the 0.234 acceptance rate for RWM is surprisingly robust even in lower dimensions across various proposal, multimodal distributions that may not have an i.i.d. product density, and curved Rosenbrock target distributions with nonlinear correlation structure. Parallel tempering experiments also show that the idealized 0.234 spacing of inverse temperatures may be approximately optimal for low dimensions and non i.i.d. product target densities, and that constructing an inverse temperature ladder with spacings given by a swap acceptance of 0.234 is a viable strategy. | [
{
"version": "v1",
"created": "Tue, 13 Aug 2024 13:42:56 GMT"
},
{
"version": "v2",
"created": "Thu, 17 Oct 2024 02:04:05 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 15:19:48 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Li",
"Aidan",
""
],
[
"Wang",
"Liyan",
""
],
[
"Dou",
"Tianye",
""
],
[
"Rosenthal",
"Jeffrey S.",
""
]
] |
2408.07535 | Otari Sakhelashvili | Gia Dvali, Archil Kobakhidze, Otari Sakhelashvili | Electroweak $\eta_{\rm w}$ meson | 14 pages, published version | Phys.Rev.D 111 (2025) 11, 113002 | 10.1103/mxdm-wb1p | null | hep-th hep-ph | http://creativecommons.org/licenses/by/4.0/ | We argue that the Standard Model is accompanied by a new pseudo-scalar degree of freedom, $\eta_{\rm w}$-meson, which cancels the topological susceptibility of the electroweak vacuum and gets its mass from this effect. The prediction is based on the analyticity properties of the Chern-Simons correlator combined with the basic features of gravity. Depending on the quality level of the $U(1)_{B+L}$-symmetry, $\eta_{\rm w}$ emerges as a $B+L$ pseudo-Goldstone boson or as a St\"uckelberg $2$-form of the electroweak gauge redundancy. An intriguing scenario of the first category is the emergence of $\eta_{\rm w}$ in the form of the phase of a $U(1)_{B+L}$-violating fermion condensate triggered by the instantons, somewhat similar to $\eta'$-meson in QCD. Regardless of its origin, the presence of $\eta_{\rm w}$-meson in the theory appears to be a matter of consistency. | [
{
"version": "v1",
"created": "Wed, 14 Aug 2024 13:10:10 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 12:10:08 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Dvali",
"Gia",
""
],
[
"Kobakhidze",
"Archil",
""
],
[
"Sakhelashvili",
"Otari",
""
]
] |
2408.07588 | Guiomar Pescador-Barrios | Guiomar Pescador-Barrios, Sarah Filippi, Mark van der Wilk | Adjusting Model Size in Continual Gaussian Processes: How Big is Big Enough? | 9 pages main, 27 pages total, 13 figures, 9 tables, conference paper, minor correction: updated author name in PDF | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many machine learning models require setting a parameter that controls their size before training, e.g. number of neurons in DNNs, or inducing points in GPs. Increasing capacity typically improves performance until all the information from the dataset is captured. After this point, computational cost keeps increasing, without improved performance. This leads to the question "How big is big enough?" We investigate this problem for Gaussian processes (single-layer neural networks) in continual learning. Here, data becomes available incrementally, and the final dataset size will therefore not be known before training, preventing the use of heuristics for setting a fixed model size. We develop a method to automatically adjust model size while maintaining near-optimal performance. Our experimental procedure follows the constraint that any hyperparameters must be set without seeing dataset properties, and we show that our method performs well across diverse datasets without the need to adjust its hyperparameter, showing it requires less tuning than others. | [
{
"version": "v1",
"created": "Wed, 14 Aug 2024 14:40:00 GMT"
},
{
"version": "v2",
"created": "Fri, 13 Dec 2024 19:36:43 GMT"
},
{
"version": "v3",
"created": "Fri, 06 Jun 2025 06:19:35 GMT"
},
{
"version": "v4",
"created": "Wed, 11 Jun 2025 07:41:25 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Pescador-Barrios",
"Guiomar",
""
],
[
"Filippi",
"Sarah",
""
],
[
"van der Wilk",
"Mark",
""
]
] |
2408.08069 | Murat Ersalman | Murat Ersalman, Mervi Kunnasranta, Markus Ahola, Anja M. Carlsson, Sara Persson, Britt-Marie B\"acklin, Inari Helle, Linnea Cervin, Jarno Vanhatalo | Integrated population model reveals human and environment driven changes in Baltic ringed seal (Pusa hispida botnica) demography and behavior | null | null | null | null | q-bio.PE stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Integrated population models (IPMs) are a promising approach to test ecological theories and assess wildlife populations in dynamic and uncertain conditions. By combining multiple data sources into a unified model, they enable the parametrization of versatile, mechanistic models that can predict population dynamics in novel circumstances. Here, we present a Bayesian IPM for the ringed seal (Pusa hispida botnica) population inhabiting the Bothnian Bay in the Baltic Sea. Despite the availability of long-term monitoring data, traditional assessment methods have faltered due to dynamic environmental conditions, varying reproductive rates, and the recently re-introduced hunting, thus limiting the quality of information available to managers. We fit our model to census and various demographic, reproductive, and harvest data from 1988 to 2023 to provide a comprehensive assessment of past population trends, and predict population response to alternative hunting scenarios. We estimated that 20,000 to 36,000 ringed seals inhabited the Bothnian Bay in 2024, increasing at a rate of 3% to 6% per year. Reproductive rates have increased since 1988, leading to a substantial increase in the growth rate up until 2015. However, the re-introduction of hunting has since reduced the growth rate, and even minor quota increases are likely to reduce it further. Our results also support the hypothesis that a greater proportion of the population hauls out under lower ice cover circumstances, leading to higher aerial survey results in such years. In general, our study demonstrates the value of IPMs for monitoring wildlife populations under changing environments, and supporting science-based management decisions. | [
{
"version": "v1",
"created": "Thu, 15 Aug 2024 10:34:23 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 16:37:10 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Ersalman",
"Murat",
""
],
[
"Kunnasranta",
"Mervi",
""
],
[
"Ahola",
"Markus",
""
],
[
"Carlsson",
"Anja M.",
""
],
[
"Persson",
"Sara",
""
],
[
"Bäcklin",
"Britt-Marie",
""
],
[
"Helle",
"Inari",
""
],
[
"Cervin",
"Linnea",
""
],
[
"Vanhatalo",
"Jarno",
""
]
] |
2408.08169 | Thomas Krainer | Thomas Krainer | A Shubin pseudodifferential calculus on asymptotically conic manifolds | This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this article is published in Journal of Fourier Analysis and Applications, and is available online at https://doi.org/10.1007/s00041-025-10178-3 | J. Fourier Anal. Appl. 31, 44 (2025) | 10.1007/s00041-025-10178-3 | MPIM-Bonn-2024 | math.AP math.FA math.OA | http://creativecommons.org/licenses/by/4.0/ | We present a global pseudodifferential calculus on asymptotically conic manifolds that generalizes (anisotropic versions of) Shubin's classical global pseudodifferential calculus on Euclidean space to this class of noncompact manifolds. Fully elliptic operators are shown to be Fredholm in an associated scale of Sobolev spaces, and to have parametrices in the calculus. | [
{
"version": "v1",
"created": "Thu, 15 Aug 2024 14:03:54 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:18:08 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Krainer",
"Thomas",
""
]
] |
2408.08171 | Emanuel Milman | Emanuel Milman, Shahar Shabelman and Amir Yehudayoff | Fixed and Periodic Points of the Intersection Body Operator | 49 pages, added figures. Final version, to appear in Invent. Math | null | null | null | math.FA math.MG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The intersection body $IK$ of a star-body $K$ in $\mathbb{R}^n$ was introduced by E. Lutwak following the work of H. Busemann, and plays a central role in the dual Brunn-Minkowski theory. We show that when $n \geq 3$, $I^2 K = c K$ iff $K$ is a centered ellipsoid, and hence $I K = c K$ iff $K$ is a centered Euclidean ball, answering long-standing questions by Lutwak, Gardner, and Fish-Nazarov-Ryabogin-Zvavitch. To this end, we recast the iterated intersection body equation as an Euler-Lagrange equation for a certain volume functional under radial perturbations, derive new formulas for the volume of $I K$, and introduce a continuous version of Steiner symmetrization for Lipschitz star-bodies, which (surprisingly) yields a useful radial perturbation exactly when $n\geq 3$. | [
{
"version": "v1",
"created": "Thu, 15 Aug 2024 14:10:53 GMT"
},
{
"version": "v2",
"created": "Thu, 22 Aug 2024 09:57:33 GMT"
},
{
"version": "v3",
"created": "Mon, 02 Sep 2024 13:31:42 GMT"
},
{
"version": "v4",
"created": "Tue, 10 Jun 2025 18:06:28 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Milman",
"Emanuel",
""
],
[
"Shabelman",
"Shahar",
""
],
[
"Yehudayoff",
"Amir",
""
]
] |
2408.08979 | Minheng Xiao | Minheng Xiao | Electroencephalogram Emotion Recognition via AUC Maximization | null | null | null | null | cs.LG eess.SP | http://creativecommons.org/licenses/by/4.0/ | Imbalanced datasets pose significant challenges in areas including neuroscience, cognitive science, and medical diagnostics, where accurately detecting minority classes is essential for robust model performance. This study addresses the issue of class imbalance, using the `Liking' label in the DEAP dataset as an example. Such imbalances are often overlooked by prior research, which typically focuses on the more balanced arousal and valence labels and predominantly uses accuracy metrics to measure model performance. To tackle this issue, we adopt numerical optimization techniques aimed at maximizing the area under the curve (AUC), thus enhancing the detection of underrepresented classes. Our approach, which begins with a linear classifier, is compared against traditional linear classifiers, including logistic regression and support vector machines (SVM). Our method significantly outperforms these models, increasing recall from 41.6\% to 79.7\% and improving the F1-score from 0.506 to 0.632. These results highlight the efficacy of AUC maximization via numerical optimization in managing imbalanced datasets, providing an effective solution for enhancing predictive accuracy in detecting minority but crucial classes in out-of-sample datasets. | [
{
"version": "v1",
"created": "Fri, 16 Aug 2024 19:08:27 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Jan 2025 02:51:13 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Jun 2025 20:55:06 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Xiao",
"Minheng",
""
]
] |
2408.09872 | Marcel Cech | Marcel Cech, Mar\'ia Cea, Mari Carmen Ba\~nuls, Igor Lesanovsky, Federico Carollo | Space-time correlations in monitored kinetically constrained discrete-time quantum dynamics | 9+8 pages, 3+4 figures, comments welcome | Phys. Rev. Lett. 134, 230403 (2025) | 10.1103/2lxs-wccj | null | quant-ph cond-mat.quant-gas cond-mat.stat-mech | http://creativecommons.org/licenses/by-sa/4.0/ | State-of-the-art quantum simulators permit local temporal control of interactions and midcircuit readout. These capabilities open the way towards the exploration of intriguing nonequilibrium phenomena. We illustrate this with a kinetically constrained many-body quantum system that has a natural implementation on Rydberg quantum simulators. The evolution proceeds in discrete time and is generated by repeatedly entangling the system with an auxiliary environment that is monitored and reset after each time-step. Despite featuring an uncorrelated infinite-temperature average stationary state, the dynamics displays coexistence of fast and slow space-time regions in stochastic realizations of the system state. The time-record of measurement outcomes on the environment serves as natural probe for such dynamical heterogeneity, which we characterize using tools from large deviation theory. Our work establishes the large deviation framework for discrete-time open quantum many-body systems as a means to characterize complex dynamics and collective phenomena in quantum processors and simulators. | [
{
"version": "v1",
"created": "Mon, 19 Aug 2024 10:24:07 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Dec 2024 09:50:55 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:07:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Cech",
"Marcel",
""
],
[
"Cea",
"María",
""
],
[
"Bañuls",
"Mari Carmen",
""
],
[
"Lesanovsky",
"Igor",
""
],
[
"Carollo",
"Federico",
""
]
] |
2408.10928 | Leonid Positselski | Leonid Positselski | A contramodule generalization of Neeman's flat and projective module theorem | LaTeX 2e, 34 pages; v.2: Remark 5.4 added, new Sections 7-10 inserted; abstract, introduction, and the last section expanded -- the cotorsion and flat cotorsion stories added to the flat and projective stories presented in the first version v1; v.3: a misprint corrected, a reference updated; Remark 8.2, with added references, inserted | null | null | null | math.RA math.CT | http://creativecommons.org/licenses/by/4.0/ | This paper builds on top of arXiv:2306.02734. We consider a complete, separated topological ring $\mathfrak R$ with a countable base of neighborhoods of zero consisting of open two-sided ideals. The main result is that the homotopy category of projective left $\mathfrak R$-contramodules is equivalent to the derived category of the exact category of flat left $\mathfrak R$-contramodules, and also to the homotopy category of flat cotorsion left $\mathfrak R$-contramodules. In other words, a complex of flat $\mathfrak R$-contramodules is contraacyclic (in the sense of Becker) if and only if it is an acyclic complex with flat $\mathfrak R$-contramodules of cocycles, and if and only if it is coacyclic as a complex in the exact category of flat $\mathfrak R$-contramodules. These are contramodule generalizations of theorems of Neeman and of Bazzoni, Cortes-Izurdiaga, and Estrada. | [
{
"version": "v1",
"created": "Tue, 20 Aug 2024 15:12:12 GMT"
},
{
"version": "v2",
"created": "Sat, 23 Nov 2024 20:15:40 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:29:25 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Positselski",
"Leonid",
""
]
] |
2408.11526 | Mayank Kharbanda | Mayank Kharbanda, Rajiv Ratn Shah, Raghava Mutharaju | RConE: Rough Cone Embedding for Multi-Hop Logical Query Answering on Multi-Modal Knowledge Graphs | Accepted in TKDE (June 2025) as regular paper | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Multi-hop query answering over a Knowledge Graph (KG) involves traversing one or more hops from the start node to answer a query. Path-based and logic-based methods are state-of-the-art for multi-hop question answering. The former is used in link prediction tasks. The latter is for answering complex logical queries. The logical multi-hop querying technique embeds the KG and queries in the same embedding space. The existing work incorporates First Order Logic (FOL) operators, such as conjunction ($\wedge$), disjunction ($\vee$), and negation ($\neg$), in queries. Though current models have most of the building blocks to execute the FOL queries, they cannot use the dense information of multi-modal entities in the case of Multi-Modal Knowledge Graphs (MMKGs). We propose RConE, an embedding method to capture the multi-modal information needed to answer a query. The model first shortlists candidate (multi-modal) entities containing the answer. It then finds the solution (sub-entities) within those entities. Several existing works tackle path-based question-answering in MMKGs. However, to our knowledge, we are the first to introduce logical constructs in querying MMKGs and to answer queries that involve sub-entities of multi-modal entities as the answer. Extensive evaluation of four publicly available MMKGs indicates that RConE outperforms the current state-of-the-art. | [
{
"version": "v1",
"created": "Wed, 21 Aug 2024 11:02:35 GMT"
},
{
"version": "v2",
"created": "Mon, 26 Aug 2024 07:46:25 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 08:52:33 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Kharbanda",
"Mayank",
""
],
[
"Shah",
"Rajiv Ratn",
""
],
[
"Mutharaju",
"Raghava",
""
]
] |
2408.12546 | Alexander Oleynichenko | Alexander V. Oleynichenko, Andrei Zaitsevskii, Leonid V. Skripnikov, Ephraim Eliav | Finite-order method to calculate approximate density matrices in the Fock-space multireference coupled cluster theory | null | Mol. Phys., Alexander Nemukhin Special Issue, e2413416 (2024) | 10.1080/00268976.2024.2413416 | null | physics.comp-ph physics.atom-ph physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An efficient approach to calculate approximate pure-state and transition reduced density matrices in the framework of the multireference relativistic Fock-space coupled cluster (FS CC) theory is proposed. The method is based on the effective operator formalism and consists of the direct substitution of the FS CC Ansatz for a wave operator into the effective operator expression with the subsequent truncation of expansion at the terms quadratic in cluster amplitudes. The final density matrix is defined by active-space density matrices of different ranks "dressed" with contributions from cluster operators. The method gives a connected expression for pure-state density matrices, provided that the intermediate normalization condition is fulfilled. Moreover, under some additional assumptions, the connectivity can also be ensured for calculated transition property matrix elements and natural transition spinors. The developed technique allows for fast and accurate calculations of one-particle reduced density matrices for a wide range of electronic states. A pilot application of the new technique to construct averaged atomic natural orbital (ANO) basis sets for fully relativistic electronic structure calculations is presented. | [
{
"version": "v1",
"created": "Thu, 22 Aug 2024 17:00:34 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Oleynichenko",
"Alexander V.",
""
],
[
"Zaitsevskii",
"Andrei",
""
],
[
"Skripnikov",
"Leonid V.",
""
],
[
"Eliav",
"Ephraim",
""
]
] |
2408.12894 | Young Sun Choi | Yunji Seo, Young Sun Choi, Hyun Seung Son, Youngjung Uh | FLoD: Integrating Flexible Level of Detail into 3D Gaussian Splatting for Customizable Rendering | Project page: https://3dgs-flod.github.io/flod/ | null | 10.1145/3731430 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | 3D Gaussian Splatting (3DGS) and its subsequent works are restricted to specific hardware setups, either on only low-cost or on only high-end configurations. Approaches aimed at reducing 3DGS memory usage enable rendering on low-cost GPU but compromise rendering quality, which fails to leverage the hardware capabilities in the case of higher-end GPU. Conversely, methods that enhance rendering quality require high-end GPU with large VRAM, making such methods impractical for lower-end devices with limited memory capacity. Consequently, 3DGS-based works generally assume a single hardware setup and lack the flexibility to adapt to varying hardware constraints.
To overcome this limitation, we propose Flexible Level of Detail (FLoD) for 3DGS. FLoD constructs a multi-level 3DGS representation through level-specific 3D scale constraints, where each level independently reconstructs the entire scene with varying detail and GPU memory usage. A level-by-level training strategy is introduced to ensure structural consistency across levels. Furthermore, the multi-level structure of FLoD allows selective rendering of image regions at different detail levels, providing additional memory-efficient rendering options. To our knowledge, among prior works which incorporate the concept of Level of Detail (LoD) with 3DGS, FLoD is the first to follow the core principle of LoD by offering adjustable options for a broad range of GPU settings.
Experiments demonstrate that FLoD provides various rendering options with trade-offs between quality and memory usage, enabling real-time rendering under diverse memory constraints. Furthermore, we show that FLoD generalizes to different 3DGS frameworks, indicating its potential for integration into future state-of-the-art developments. | [
{
"version": "v1",
"created": "Fri, 23 Aug 2024 07:56:25 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 04:18:49 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Seo",
"Yunji",
""
],
[
"Choi",
"Young Sun",
""
],
[
"Son",
"Hyun Seung",
""
],
[
"Uh",
"Youngjung",
""
]
] |
2408.14229 | Leonid Erlygin | Leonid Erlygin, Alexey Zaytsev | Holistic Uncertainty Estimation For Open-Set Recognition | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate uncertainty estimation is a critical challenge in open-set recognition, where a probe biometric sample may belong to an unknown identity. It can be addressed through sample quality estimation via probabilistic embeddings. However, the low variance of probabilistic embedding only partly implies a low identification error probability: an embedding of a sample could be close to several classes in a gallery, thus yielding high uncertainty despite high sample quality. We propose HolUE - a holistic uncertainty estimation method based on a Bayesian probabilistic model; it is aware of two sources of ambiguity in the open-set recognition system: (1) the gallery uncertainty caused by overlapping classes and (2) the uncertainty of embeddings. Challenging open-set recognition datasets, such as IJB-C for the image domain and VoxBlink for the audio domain, serve as a testbed for our method. We also provide a new open-set recognition protocol for the identification of whales and dolphins. In all cases, HolUE better identifies recognition errors than alternative uncertainty estimation methods, including those based solely on sample quality. | [
{
"version": "v1",
"created": "Mon, 26 Aug 2024 12:44:17 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 11:35:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Erlygin",
"Leonid",
""
],
[
"Zaytsev",
"Alexey",
""
]
] |
2408.14352 | Nicolas Yax | Nicolas Yax and Pierre-Yves Oudeyer and Stefano Palminteri | LogProber: Disentangling confidence from contamination in LLM responses | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | In machine learning, contamination refers to situations where testing data leak into the training set. The issue is particularly relevant for the evaluation of the performance of Large Language Models (LLMs), which are generally trained on gargantuan, and generally opaque, corpora of text scraped from the world wide web. Developing tools to detect contamination is therefore crucial to be able to fairly and properly track the evolution of the performance of LLMs. To date, only a few recent studies have attempted to address the issue of quantifying and detecting contamination in short text sequences, such as those commonly found in benchmarks. However, these methods have limitations that can sometimes render them impractical.In the present paper, we introduce LogProber, a novel, efficient algorithm that we show to be able to detect contamination in a black box setting that tries to tackle some of these drawbacks by focusing on the familiarity with the question rather than the answer. Here, we explore the properties of the proposed method in comparison with concurrent approaches, identify its advantages and limitations, and illustrate how different forms of contamination can go undetected depending on the design of the detection algorithm. | [
{
"version": "v1",
"created": "Mon, 26 Aug 2024 15:29:34 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:42:54 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Yax",
"Nicolas",
""
],
[
"Oudeyer",
"Pierre-Yves",
""
],
[
"Palminteri",
"Stefano",
""
]
] |
2408.15028 | Simon Warder | Simon C Warder and Matthew D Piggott | Mapping global offshore wind wake losses, layout optimisation potential, and climate change effects | null | Energy, 331, p.136573 (2025) | 10.1016/j.energy.2025.136573 | null | physics.ao-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | This study assesses global offshore wind energy resources, wake-induced losses, array layout optimisation potential and climate change impacts. Global offshore ambient potential is first mapped based on reanalysis data. Wake-induced losses are then estimated using an engineering wake model, revealing that locations with low (high) resource typically experience larger (smaller) percentage losses. However, the specific wind speed distribution is found to be important, with narrower distributions generally leading to greater losses. This is due to the overlap between the wind speed distribution and the high-sensitivity region of the turbine thrust and power curves. Broadly, this leads to much stronger wake-induced losses in the tropics (which experience the trade winds) than mid-latitudes. However, the tropics also experience a narrower wind direction distribution; the results of this study demonstrate that this leads to greater potential for mitigation of wake effects via layout optimisation. Finally, projected changes in wind potential and wake losses due to climate change under a high-emission scenario are assessed. Many regions are projected to decrease in ambient wind resources, and furthermore these regions will typically experience greater wake-induced losses, exacerbating the climate change impact. These results highlight the different challenges and opportunities associated with exploiting offshore wind resources across the globe. | [
{
"version": "v1",
"created": "Tue, 27 Aug 2024 12:57:17 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jun 2025 14:37:59 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Warder",
"Simon C",
""
],
[
"Piggott",
"Matthew D",
""
]
] |
2408.16022 | Jeremy Wayland | Jeremy Wayland, Russel J. Funk, Bastian Rieck | Characterizing Physician Referral Networks with Ricci Curvature | null | Communications in Computer and Information Science, 2386:3-17, 2025 | 10.1007/978-3-031-88346-0_1 | null | cs.SI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Identifying (a) systemic barriers to quality healthcare access and (b) key indicators of care efficacy in the United States remains a significant challenge. To improve our understanding of regional disparities in care delivery, we introduce a novel application of curvature, a geometrical-topological property of networks, to Physician Referral Networks. Our initial findings reveal that Forman-Ricci and Ollivier-Ricci curvature measures, which are known for their expressive power in characterizing network structure, offer promising indicators for detecting variations in healthcare efficacy while capturing a range of significant regional demographic features. We also present APPARENT, an open-source tool that leverages Ricci curvature and other network features to examine correlations between regional Physician Referral Networks structure, local census data, healthcare effectiveness, and patient outcomes. | [
{
"version": "v1",
"created": "Tue, 27 Aug 2024 08:20:35 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Oct 2024 09:39:20 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Wayland",
"Jeremy",
""
],
[
"Funk",
"Russel J.",
""
],
[
"Rieck",
"Bastian",
""
]
] |
2408.16326 | Xin Zheng | Xin Zheng, Jie Lou, Boxi Cao, Xueru Wen, Yuqiu Ji, Hongyu Lin, Yaojie Lu, Xianpei Han, Debing Zhang, Le Sun | Critic-CoT: Boosting the reasoning abilities of large language model via Chain-of-thoughts Critic | Accepted at ACL 2025 Findings | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-critic has become a crucial mechanism for enhancing the reasoning performance of LLMs. However, current approaches mainly involve basic prompts for intuitive instance-level feedback, which resembles System-1 processes and limits the reasoning capabilities. Moreover, there is a lack of in-depth investigations into the relationship between LLM's ability to criticize and its task-solving performance. To address these issues, we propose Critic-CoT, a novel framework that pushes LLMs toward System-2-like critic capability. Through a step-wise CoT reasoning paradigm and the automatic construction of distant-supervision data without human annotation, Critic-CoT enables LLMs to engage in slow, analytic self-critique and refinement, thereby improving their reasoning abilities. Experiments on GSM8K and MATH demonstrate that our enhanced model significantly boosts task-solving performance by filtering out invalid solutions or iterative refinement. Furthermore, we investigate the intrinsic correlation between critique and task-solving abilities within LLMs, discovering that these abilities can mutually reinforce each other rather than conflict. | [
{
"version": "v1",
"created": "Thu, 29 Aug 2024 08:02:09 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Oct 2024 06:42:16 GMT"
},
{
"version": "v3",
"created": "Wed, 11 Jun 2025 02:01:02 GMT"
}
] | 2025-06-12T00:00:00 | [
[
"Zheng",
"Xin",
""
],
[
"Lou",
"Jie",
""
],
[
"Cao",
"Boxi",
""
],
[
"Wen",
"Xueru",
""
],
[
"Ji",
"Yuqiu",
""
],
[
"Lin",
"Hongyu",
""
],
[
"Lu",
"Yaojie",
""
],
[
"Han",
"Xianpei",
""
],
[
"Zhang",
"Debing",
""
],
[
"Sun",
"Le",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.