id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2507.01601 | Constance Mahony | Constance Mahony, Sergio Contreras, Raul E. Angulo, David Alonso, Christos Georgiou, Andrej Dvornik | Cosmological constraints from galaxy clustering and galaxy-galaxy lensing with extended SubHalo Abundance Matching | 13 pages, 9 figures, submitted to MNRAS | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | We present the first cosmological constraints from a joint analysis of galaxy clustering and galaxy-galaxy lensing using extended SubHalo Abundance Matching (SHAMe). We analyse stellar mass-selected Galaxy And Mass Assembly (GAMA) galaxy clustering and Kilo-Degree Survey (KiDS-1000) galaxy-galaxy lensing and find constraints on $S_8\equiv\sigma_8\sqrt{\Omega_{\rm m}/0.3}=0.793^{+0.025}_{-0.024}$, in agreement with Planck at 1.7$\sigma$, and consistent with previous results. We are able to constrain all 5 SHAMe parameters, which describe the galaxy-subhalo connection. We validate our methodology by first applying it to simulated catalogues, generated from the TNG300 simulation, which mimic the stellar mass selection of our real data. We show that we are able to recover the input cosmology for both our fiducial and all-scale analyses. Our all-scale analysis extends to scales of galaxy-galaxy lensing below $r_\mathrm{p}<1.4\,\mathrm{Mpc}/h$, which we exclude in our fiducial analysis to avoid baryonic effects. When including all scales, we find a value of $S_8$, which is 1.26$\sigma$ higher than our fiducial result (against naive expectations where baryonic feedback should lead to small-scale power suppression), and in agreement with Planck at 0.9$\sigma$. We also find a 21% tighter constraint on $S_8$ and a 29% tighter constraint on $\Omega_\mathrm{m}$ compared to our fiducial analysis. This work shows the power and potential of joint small-scale galaxy clustering and galaxy-galaxy lensing analyses using SHAMe. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:09:34 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mahony",
"Constance",
""
],
[
"Contreras",
"Sergio",
""
],
[
"Angulo",
"Raul E.",
""
],
[
"Alonso",
"David",
""
],
[
"Georgiou",
"Christos",
""
],
[
"Dvornik",
"Andrej",
""
]
] |
2507.01602 | Kun Zhang | Kun Zhang, Mo-Yang Ni, Hai-Long Shi, Xiao-Hui Wang, Jin Wang | Many-Body Fluctuation Theorems for Quantum Coherence and Correlation Dynamics | 8 pages, 3 figures | null | null | null | quant-ph cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fluctuation theorems establish exact relations for nonequilibrium dynamics, profoundly advancing the field of stochastic thermodynamics. In this Letter, we extend quantum fluctuation theorems beyond the traditional thermodynamic framework to quantum information dynamics and many-body systems, where both the system and the environment are multipartite without assuming any thermodynamic constraints. Based on the two-point measurement scheme and the classical probability, we establish the fluctuation theorem for the dynamics of many-body classical mutual information. By extending to quasiprobability, we derive quantum fluctuation theorems for many-body coherence and quantum correlations, presenting them in both integral and detailed forms. Our theoretical results are illustrated and verified using three-qubit examples, and feasible experimental verification protocols are proposed. These findings uncover the statistical structure underlying the nonequilibrium quantum information dynamics, providing fundamental insights and new tools for advancing quantum technologies. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:10:42 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhang",
"Kun",
""
],
[
"Ni",
"Mo-Yang",
""
],
[
"Shi",
"Hai-Long",
""
],
[
"Wang",
"Xiao-Hui",
""
],
[
"Wang",
"Jin",
""
]
] |
2507.01603 | Yuejiang Dong | Yue-Jiang Dong, Wang Zhao, Jiale Xu, Ying Shan, Song-Hai Zhang | DepthSync: Diffusion Guidance-Based Depth Synchronization for Scale- and Geometry-Consistent Video Depth Estimation | Accepted by ICCV 2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion-based video depth estimation methods have achieved remarkable success with strong generalization ability. However, predicting depth for long videos remains challenging. Existing methods typically split videos into overlapping sliding windows, leading to accumulated scale discrepancies across different windows, particularly as the number of windows increases. Additionally, these methods rely solely on 2D diffusion priors, overlooking the inherent 3D geometric structure of video depths, which results in geometrically inconsistent predictions. In this paper, we propose DepthSync, a novel, training-free framework using diffusion guidance to achieve scale- and geometry-consistent depth predictions for long videos. Specifically, we introduce scale guidance to synchronize the depth scale across windows and geometry guidance to enforce geometric alignment within windows based on the inherent 3D constraints in video depths. These two terms work synergistically, steering the denoising process toward consistent depth predictions. Experiments on various datasets validate the effectiveness of our method in producing depth estimates with improved scale and geometry consistency, particularly for long videos. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:11:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Dong",
"Yue-Jiang",
""
],
[
"Zhao",
"Wang",
""
],
[
"Xu",
"Jiale",
""
],
[
"Shan",
"Ying",
""
],
[
"Zhang",
"Song-Hai",
""
]
] |
2507.01604 | D\'aniel N\'emeth Dr. | Timothy Budd, D\'aniel N\'emeth | Phases of Tree-decorated Dynamical Triangulations in 3D | null | null | null | null | hep-lat gr-qc hep-th math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | This work revisits the Euclidean Dynamical Triangulation (DT) approach to non-perturbative quantum gravity in three dimensions. Inspired by a recent combinatorial study by T. Budd and L. Lionni of a subclass of 3-sphere triangulations constructed from trees, called the \emph{triple-tree} class, we present a Monte Carlo investigation of DT decorated with a pair of spanning trees, one spanning the vertices and the other the tetrahedra of the triangulation. The complement of the pair of trees in the triangulation can be viewed as a bipartite graph, called the \emph{middle graph} of the triangulation. In the triple-tree class, the middle graph is restricted to be a tree, and numerical simulations have displayed a qualitatively different phase structure compared to standard DT. Relaxing this restriction, the middle graph comes with two natural invariants, namely the number of connected components and loops. Introducing corresponding coupling constants in the action, allows one to interpolate between the triple-tree class and unrestricted tree-decorated DT. Simulations of this extended model confirm the existence of a new phase, referred to as the \emph{triple-tree phase}, besides the familiar crumpled and branched polymer phases of DT. A statistical analysis of the phase transitions is presented, showing hints that the branched polymer to triple-tree phase transition is continuous. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:11:58 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Budd",
"Timothy",
""
],
[
"Németh",
"Dániel",
""
]
] |
2507.01605 | G\'abor Homa | G\'abor Homa, D\'avid Hamar, J\'ozsef Zsolt Bern\'ad, Peter Adam, Andr\'as Csord\'as | An exact Markovian decoherence dynamics of two interacting harmonic oscillators coupled to a bosonic heat bath | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present an exact analytical solution of the Hu-Paz-Zhang master equation in a precise Markovian limit for a system of two harmonically coupled harmonic oscillators interacting with a common thermal bath of harmonic oscillators. The thermal bath is initially considered to be at arbitrary temperatures and characterized by an Ohmic Lorentz-Drude spectral density. In the examined system, couplings between the two harmonic oscillators and the environment ensure a complete decoupling of the center-of-mass and relative degrees of freedom, resulting in undamped dynamics in the relative coordinate. The exact time evolution is used to analyze the system's entanglement dynamics, quantified through logarithmic negativity and quantum mutual information, while ensuring the positivity of the density operator to confirm the physical validity of the results. We demonstrate that, under certain parameter regimes and initial conditions, the asymptotic dynamics can give rise to periodic entanglement-disentanglement behavior. Furthermore, numerical simulations reveal that for negative values of the direct coupling between the oscillators, which are sufficiently close to a critical lower bound beyond which the system becomes unstable, the system can maintain entanglement across a broad temperature range and for arbitrarily long durations. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:13:26 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Homa",
"Gábor",
""
],
[
"Hamar",
"Dávid",
""
],
[
"Bernád",
"József Zsolt",
""
],
[
"Adam",
"Peter",
""
],
[
"Csordás",
"András",
""
]
] |
2507.01606 | Nick Van Loo | Nick van Loo, Francesco Zatelli, Gorm O. Steffensen, Bart Roovers, Guanzhong Wang, Thomas Van Caekenberghe, Alberto Bordin, David van Driel, Yining Zhang, Wietze D. Huisman, Ghada Badawy, Erik P.A.M. Bakkers, Grzegorz P. Mazur, Ram\'on Aguado, Leo P. Kouwenhoven | Single-shot parity readout of a minimal Kitaev chain | null | null | null | null | cond-mat.mes-hall cond-mat.supr-con | http://creativecommons.org/licenses/by/4.0/ | Protecting qubits from noise is essential for building reliable quantum computers. Topological qubits offer a route to this goal by encoding quantum information non-locally, using pairs of Majorana zero modes. These modes form a shared fermionic state whose occupation -- either even or odd -- defines the fermionic parity that encodes the qubit. Crucially, this parity cannot be accessed by any measurement that probes only one Majorana mode. This reflects the non-local nature of the encoding and its inherent protection against noise. A promising platform for realizing such qubits is the Kitaev chain, implemented in quantum dots coupled via superconductors. Even a minimal chain of two dots can host a pair of Majorana modes and store quantum information in their joint parity. Here we introduce a new technique for reading out this parity, based on quantum capacitance. This global probe senses the joint state of the chain and enables real-time, single-shot discrimination of the parity state. By comparing with simultaneous local charge sensing, we confirm that only the global signal resolves the parity. We observe random telegraph switching and extract parity lifetimes exceeding one millisecond. These results establish the essential readout step for time-domain control of Majorana qubits, resolving a long-standing experimental challenge. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:14:54 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"van Loo",
"Nick",
""
],
[
"Zatelli",
"Francesco",
""
],
[
"Steffensen",
"Gorm O.",
""
],
[
"Roovers",
"Bart",
""
],
[
"Wang",
"Guanzhong",
""
],
[
"Van Caekenberghe",
"Thomas",
""
],
[
"Bordin",
"Alberto",
""
],
[
"van Driel",
"David",
""
],
[
"Zhang",
"Yining",
""
],
[
"Huisman",
"Wietze D.",
""
],
[
"Badawy",
"Ghada",
""
],
[
"Bakkers",
"Erik P. A. M.",
""
],
[
"Mazur",
"Grzegorz P.",
""
],
[
"Aguado",
"Ramón",
""
],
[
"Kouwenhoven",
"Leo P.",
""
]
] |
2507.01607 | Yannick Teglia | Quentin Le Roux, Yannick Teglia, Teddy Furon, Philippe Loubet-Moundi, Eric Bourbao | Survivability of Backdoor Attacks on Unconstrained Face Recognition Systems | null | null | null | null | cs.CV cs.AI cs.CR cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The widespread use of deep learning face recognition raises several security concerns. Although prior works point at existing vulnerabilities, DNN backdoor attacks against real-life, unconstrained systems dealing with images captured in the wild remain a blind spot of the literature. This paper conducts the first system-level study of backdoors in deep learning-based face recognition systems. This paper yields four contributions by exploring the feasibility of DNN backdoors on these pipelines in a holistic fashion. We demonstrate for the first time two backdoor attacks on the face detection task: face generation and face landmark shift attacks. We then show that face feature extractors trained with large margin losses also fall victim to backdoor attacks. Combining our models, we then show using 20 possible pipeline configurations and 15 attack cases that a single backdoor enables an attacker to bypass the entire function of a system. Finally, we provide stakeholders with several best practices and countermeasures. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:21:27 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Roux",
"Quentin Le",
""
],
[
"Teglia",
"Yannick",
""
],
[
"Furon",
"Teddy",
""
],
[
"Loubet-Moundi",
"Philippe",
""
],
[
"Bourbao",
"Eric",
""
]
] |
2507.01608 | Xu Zhang | Xu Zhang, Ming Lu, Yan Chen, Zhan Ma | Perception-Oriented Latent Coding for High-Performance Compressed Domain Semantic Inference | International Conference on Multimedia and Expo (ICME), 2025 | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, compressed domain semantic inference has primarily relied on learned image coding models optimized for mean squared error (MSE). However, MSE-oriented optimization tends to yield latent spaces with limited semantic richness, which hinders effective semantic inference in downstream tasks. Moreover, achieving high performance with these models often requires fine-tuning the entire vision model, which is computationally intensive, especially for large models. To address these problems, we introduce Perception-Oriented Latent Coding (POLC), an approach that enriches the semantic content of latent features for high-performance compressed domain semantic inference. With the semantically rich latent space, POLC requires only a plug-and-play adapter for fine-tuning, significantly reducing the parameter count compared to previous MSE-oriented methods. Experimental results demonstrate that POLC achieves rate-perception performance comparable to state-of-the-art generative image coding methods while markedly enhancing performance in vision tasks, with minimal fine-tuning overhead. Code is available at https://github.com/NJUVISION/POLC. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:21:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhang",
"Xu",
""
],
[
"Lu",
"Ming",
""
],
[
"Chen",
"Yan",
""
],
[
"Ma",
"Zhan",
""
]
] |
2507.01609 | Sugumi Kanno | Taiki Ikeda, Youka Kaku, Sugumi Kanno, Jiro Soda | Toward graviton detection via photon-graviton quantum state conversion | 20 pages, 1 figure | null | null | KOBE-COSMO-25-12 | quant-ph gr-qc hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A magnetic field enables the interconversion of photons and gravitons, yet the process is usually analysed only at the level of classical wave equations. We revisit photon-graviton conversion in a quantum field theoretic framework, allowing us to track the evolution of arbitrary quantum states. Treating the photons as squeezed coherent states and the gravitons as the squeezed vacuum expected for primordial gravitational waves, we derive analytic expressions for the conversion probability and show that it can be significantly enhanced compared to the conventional estimate. We further demonstrate that the conversion both swaps preexisting entanglement and generates genuinely new entanglement between the electromagnetic and gravitational sectors, which is impossible in any classical description. Detecting such nonclassical correlations would constitute compelling evidence for the quantization of gravity and offers a novel pathway toward graviton detection. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:23:32 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ikeda",
"Taiki",
""
],
[
"Kaku",
"Youka",
""
],
[
"Kanno",
"Sugumi",
""
],
[
"Soda",
"Jiro",
""
]
] |
2507.01610 | Bilal Farooq | Farzan Moosavi and Bilal Farooq | Sphereabout: A Spherical Intersection Design for Networked Urban Air Mobility | Proceedings of 2025 IEEE Intelligent Transportation Systems Conference | null | null | null | math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Urban aerial mobility is rapidly expanding, specifically on-demand Unmanned Aerial Vehicle (UAV) delivery services in urban environments. This necessitates management of the low-altitude airspace network to ensure smooth and safe traffic throughput. This study introduces a novel three-dimensional aerial network design inspired by a terrestrial transportation graph network for UAV networked mobility. Utilizing two-way tube corridors for node-to-node delivery and a spherical roundabout model, as "Sphereabout", to optimize the traffic flow through the spherical intersection. Through this architecture, three-dimensional conflict-free air mobility management can be achieved via numerical experiments and utilizing the geometrical features of this intersection. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:25:08 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Moosavi",
"Farzan",
""
],
[
"Farooq",
"Bilal",
""
]
] |
2507.01611 | Shaowen Chen | Shaowen Chen and Tomoki Toda | QHARMA-GAN: Quasi-Harmonic Neural Vocoder based on Autoregressive Moving Average Model | This manuscript is currently under review for publication in the IEEE Transactions on Audio, Speech, and Language Processing. This work has been submitted to the IEEE for possible publication | null | null | null | eess.AS cs.SD eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vocoders, encoding speech signals into acoustic features and allowing for speech signal reconstruction from them, have been studied for decades. Recently, the rise of deep learning has particularly driven the development of neural vocoders to generate high-quality speech signals. On the other hand, the existing end-to-end neural vocoders suffer from a black-box nature that blinds the speech production mechanism and the intrinsic structure of speech, resulting in the ambiguity of separately modeling source excitation and resonance characteristics and the loss of flexibly synthesizing or modifying speech with high quality. Moreover, their sequence-wise waveform generation usually requires complicated networks, leading to substantial time consumption. In this work, inspired by the quasi-harmonic model (QHM) that represents speech as sparse components, we combine the neural network and QHM synthesis process to propose a novel framework for the neural vocoder. Accordingly, speech signals can be encoded into autoregressive moving average (ARMA) functions to model the resonance characteristics, yielding accurate estimates of the amplitudes and phases of quasi-harmonics at any frequency. Subsequently, the speech can be resynthesized and arbitrarily modified in terms of pitch shifting and time stretching with high quality, whereas the time consumption and network size decrease. The experiments indicate that the proposed method leverages the strengths of QHM, the ARMA model, and neural networks, leading to the outperformance of our methods over other methods in terms of generation speed, synthesis quality, and modification flexibility. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:28:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Chen",
"Shaowen",
""
],
[
"Toda",
"Tomoki",
""
]
] |
2507.01612 | Olli Sipil\"a | O. Sipil\"a, K. Silsbee, N. Carbajal, P. Caselli, M. Padovani | The Effect of Weak Cosmic Ray Heating Events on the Desorption of $\rm H_2$ | Accepted for publication in ApJ | null | null | null | astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The typical amount of molecular hydrogen (${\rm H_2}$) in interstellar ices is not known, but significant freeze-out of ${\rm H_2}$ on dust grains is not expected. However, chemical models ubiquitously predict large amounts of $\rm H_2$ freeze-out in dense cloud conditions, and specialized treatments are needed to control the $\rm H_2$ population on grains. Here we present a numerical desorption model where the effect of weak heating events induced by cosmic rays (CRs) that heat grains to temperatures of a few tens of Kelvin at high frequencies is included, improving upon earlier desorption models that only consider strong heating events (maximum grain temperature close to 100 K) that occur at a low frequency. A temperature of a few tens of Kelvin is high enough to induce efficient desorption of $\rm H_2$, but we find that even the weak heating events do not occur often enough to lead to significant $\rm H_2$ desorption. Taking the weak heating events into account does affect the predicted abundances of other lightly-bound species, but the effect is restricted to low column densities. We make here the canonical assumption that the grains are spherical with a radius of 0.1 $\mu$m. It is conceivable that in the case of a grain size distribution, weak heating events could provide a boost to $\rm H_2$ desorption coming off small grains, which are the most numerous. Further studies are still required to better quantify the role of CRs in the desorption of $\rm H_2$ and other weakly bound species. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:30:10 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Sipilä",
"O.",
""
],
[
"Silsbee",
"K.",
""
],
[
"Carbajal",
"N.",
""
],
[
"Caselli",
"P.",
""
],
[
"Padovani",
"M.",
""
]
] |
2507.01613 | Shirong Xu | Shirong Xu, Jingnan Zhang, Junhui Wang | When Less Is More: Binary Feedback Can Outperform Ordinal Comparisons in Ranking Recovery | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Paired comparison data, where users evaluate items in pairs, play a central role in ranking and preference learning tasks. While ordinal comparison data intuitively offer richer information than binary comparisons, this paper challenges that conventional wisdom. We propose a general parametric framework for modeling ordinal paired comparisons without ties. The model adopts a generalized additive structure, featuring a link function that quantifies the preference difference between two items and a pattern function that governs the distribution over ordinal response levels. This framework encompasses classical binary comparison models as special cases, by treating binary responses as binarized versions of ordinal data. Within this framework, we show that binarizing ordinal data can significantly improve the accuracy of ranking recovery. Specifically, we prove that under the counting algorithm, the ranking error associated with binary comparisons exhibits a faster exponential convergence rate than that of ordinal data. Furthermore, we characterize a substantial performance gap between binary and ordinal data in terms of a signal-to-noise ratio (SNR) determined by the pattern function. We identify the pattern function that minimizes the SNR and maximizes the benefit of binarization. Extensive simulations and a real application on the MovieLens dataset further corroborate our theoretical findings. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:30:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Xu",
"Shirong",
""
],
[
"Zhang",
"Jingnan",
""
],
[
"Wang",
"Junhui",
""
]
] |
2507.01614 | Hassan Alhussein | H. Alhussein | Gr\"obner-Shirshov bases of Rota-Baxter algebra of weight $\lambda$ with spectrum lying in $\{0,-\lambda\}$ | arXiv admin note: text overlap with arXiv:1903.02960 by other authors | null | null | null | math.RA | http://creativecommons.org/licenses/by/4.0/ | It is known that if $A$ is a finite-dimensional unital algebra equipped with a Rota-Baxter operator $R$ of weight $\lambda$, then spectrum of $R$ is a subset of $\{0,-\lambda\}$. We are interested on finding all consequences of the Rota-Baxter relation and the relation of the form $R^k(R+\lambda{\rm id})^l = 0$. In 2024, H.~Qiu, S. Zheng, Y. Dan solved this problem for $k = l = 1$ and $\lambda\neq0$. We find a Gr\"obner-Shirshov basis of the ideal generated by these two relations in general case. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:30:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Alhussein",
"H.",
""
]
] |
2507.01616 | Xiangmin Zhou | Chengkun He, Xiangmin Zhou, Chen Wang, Longbing Cao, Jie Shao, Xiaodong Li, Guang Xu, Carrie Jinqiu Hu, Zahir Tari | Enhanced Influence-aware Group Recommendation for Online Media Propagation | null | null | null | null | cs.IR cs.AI cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Group recommendation over social media streams has attracted significant attention due to its wide applications in domains such as e-commerce, entertainment, and online news broadcasting. By leveraging social connections and group behaviours, group recommendation (GR) aims to provide more accurate and engaging content to a set of users rather than individuals. Recently, influence-aware GR has emerged as a promising direction, as it considers the impact of social influence on group decision-making. In earlier work, we proposed Influence-aware Group Recommendation (IGR) to solve this task. However, this task remains challenging due to three key factors: the large and ever-growing scale of social graphs, the inherently dynamic nature of influence propagation within user groups, and the high computational overhead of real-time group-item matching.
To tackle these issues, we propose an Enhanced Influence-aware Group Recommendation (EIGR) framework. First, we introduce a Graph Extraction-based Sampling (GES) strategy to minimise redundancy across multiple temporal social graphs and effectively capture the evolving dynamics of both groups and items. Second, we design a novel DYnamic Independent Cascade (DYIC) model to predict how influence propagates over time across social items and user groups. Finally, we develop a two-level hash-based User Group Index (UG-Index) to efficiently organise user groups and enable real-time recommendation generation. Extensive experiments on real-world datasets demonstrate that our proposed framework, EIGR, consistently outperforms state-of-the-art baselines in both effectiveness and efficiency. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:34:17 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"He",
"Chengkun",
""
],
[
"Zhou",
"Xiangmin",
""
],
[
"Wang",
"Chen",
""
],
[
"Cao",
"Longbing",
""
],
[
"Shao",
"Jie",
""
],
[
"Li",
"Xiaodong",
""
],
[
"Xu",
"Guang",
""
],
[
"Hu",
"Carrie Jinqiu",
""
],
[
"Tari",
"Zahir",
""
]
] |
2507.01617 | Mohammed Al Kobaisi Dr. | Faisal Aljaberi, Hadi Belhaj, Sajjad Foroughi, Mohammed Al-Kobaisi, Martin Blunt | Spatially Distributed Wettability Characterization in Porous Media | null | null | null | null | cs.CE | http://creativecommons.org/licenses/by/4.0/ | An enhanced geometric algorithm for automated pore-by-pore contact angle measurement from micro-CT images, is presented that achieves superior accuracy compared to existing methods through robust fluid-fluid and solid-fluid interface extrapolation. Using this high resolution data, we generate spatially distributed contact angle maps that reveal previously hidden wettability heterogeneity. Our analysis of mixed-wet systems demonstrates the severe limitations of averaged metrics: a sample with a mean contact angle of 64.7 degrees, conventionally classified as uniformly weakly water-wet, exhibits 40% of its pore space in the intermediate-wetting regime (70-110 degrees). This heterogeneity explains the presence of minimal surface interfaces and fundamentally different pore-filling mechanisms operating within the same sample. By providing open-source tools for spatially-resolved wettability characterization, this work enables more accurate predictions of multiphase flow behavior in heterogeneous porous materials, essential for optimizing subsurface energy storage and recovery processes. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:37:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Aljaberi",
"Faisal",
""
],
[
"Belhaj",
"Hadi",
""
],
[
"Foroughi",
"Sajjad",
""
],
[
"Al-Kobaisi",
"Mohammed",
""
],
[
"Blunt",
"Martin",
""
]
] |
2507.01618 | Patrik Knopf | Patrik Knopf and Yadong Liu | A Thermodynamically Consistent Free Boundary Model for Two-Phase Flows in an Evolving Domain with Moving Contact Lines, Variable Contact Angles and Bulk-Surface Interaction | null | null | null | null | math.AP math-ph math.MP physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | We derive a thermodynamically consistent model, which describes the time evolution of a two-phase flow in an evolving domain. The movement of the free boundary of the domain is driven by the velocity field of the mixture in the bulk, which is determined by a Navier--Stokes equation. In order to take interactions between bulk and boundary into account, we further consider two materials on the boundary, which may be the same or different materials as those in the bulk. The bulk and the surface materials are represented by respective phase-fields, whose time evolution is described by a bulk-surface convective Cahn--Hilliard equation. This approach allows for a transfer of material between bulk and surface as well as variable contact angles between the diffuse interface in the bulk and the boundary of the domain. To provide a more accurate description of the corresponding contact line motion, we include a generalized Navier slip boundary condition on the velocity field. We derive our model from scratch by considering local mass balance and energy dissipation laws. Finally, the derivation is completed via the Langrange multiplier approach. We further show that our model generalizes previous models from the literature, which can be recovered from our system by either dropping the dynamic boundary conditions or assuming a static boundary of the domain. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:38:07 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Knopf",
"Patrik",
""
],
[
"Liu",
"Yadong",
""
]
] |
2507.01619 | Gregory Ben | Gregory Mathews Ben, Jessy Jose, Jes\'us Hern\'andez | Protoplanetary Disk Survival Time-scales: A Blind Survey of Young Clusters up to 100 Myr in the Solar Vicinity | Accepted for publication in MNRAS. The paper has seventeen pages, twelve figures, and two tables | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | We study the protoplanetary disk lifetimes using a large sample of young stellar objects in nearby clusters. To investigate the final phase of disk dissipation, we selected 32 clusters, located within 500 pc and aged between 1 and 100 Myr, with membership determined using Gaia data. The Age and mass information of the sources are obtained through spectral energy distribution (SED) analysis and using evolutionary models of various ages. Using the IR data from 2MASS and WISE catalogues, we employ three methods to identify disks across the different wavelength regimes (1.1- 22 $\mu$m). We find that disk fraction consistently decreases as stellar systems age, a trend observed across all wavelengths included in this study. However, there is an increase in the time scale of disk decay as wavelength increases, with characteristic timescales of $\tau_{\text{short}}$ = 1.6 $\pm$ 0.1 Myr for shorter wavelengths (1.6-4.6 $\mu$m) versus $\tau_{\text{W3}}$ = 4.4 $\pm$ 0.3 Myr for 12 $\mu$m. This supports the idea that outer disk regions evolve more slowly. Notably, we detect infrared excesses at 12 $\mu$m and 22 $\mu$m in relatively older systems ($>$10 Myr), with some disks with estimated ages up to $\sim$ 100 Myr. Among these, we identify a population of full disks that persist beyond the typical dissipation timescale. We also observe that the median mass of disk-hosting stars decreases from 0.62 $M_\odot$ to 0.27 $M_\odot$ in clusters younger and older than 40 Myr, respectively, indicating slower disk dissipation around lower-mass stars. We identify 33 transitional disk candidates using various color-color diagrams. Using LAMOST DR8 optical spectra and H-alpha equivalent widths, we identify possible accretors and estimate their mass accretion rates, finding most are younger than 10 Myr. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:44:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ben",
"Gregory Mathews",
""
],
[
"Jose",
"Jessy",
""
],
[
"Hernández",
"Jesús",
""
]
] |
2507.01620 | Nahyeon Lee | Nahyeon Lee, Jongsoo Lim, and Hyeong-Chai Jeong | Patterns and Dynamics of Netflix TV Show Popularity | 20 pages, 7 figures, 1 table (in appendix) | null | null | null | physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the global dynamics of media consumption by analyzing daily top-ranked Netflix TV shows across 71 countries over a span of 822 days. Using an information-theoretic framework, we quantify diversity, similarity, and directional influence in consumption trends through three core measures, Shannon entropy, mutual information, and Kullback-Leibler (KL) divergence. Entropy reveals that content diversity varies by region, with North America and Europe exhibiting highly dynamic viewing preferences, while East and Southeast Asian (ESA) countries maintain more stable and persistent trends. Mutual information identifies clear regional clusters of synchronized consumption, with particularly strong alignment among ESA countries. To analyze temporal flow, we introduce a KL-based asymmetry measure that captures directional influence between countries and applies to both closely connected regional groups and loosely connected cross-regional pairs. This analysis reveals distinct pathways of content diffusion. For inter-group diffusion, we observe flows from ESA and South America toward North America and Europe. For intra-group diffusion, we find influence from Korea and Thailand to other ESA countries. We also observe that ESA trends reaching other regions often originate from Singapore and the Philippines. These findings offer insight into the temporal structure of global cultural diffusion and highlight the coexistence of global synchronization and regional independence in streaming media preferences. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:45:01 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Lee",
"Nahyeon",
""
],
[
"Lim",
"Jongsoo",
""
],
[
"Jeong",
"Hyeong-Chai",
""
]
] |
2507.01621 | Alicia Mascare\~nas | Alicia Mascare\~nas-Pazos, Silvia Lorenzo-Freire, Jose Maria Alonso-Meijide | Enriching the Felsenthal index with a priori unions for decision-making processes | null | null | null | null | econ.TH cs.GT math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Within the domain of game theory, power indexes are defined as functions that quantify the influence of individual participants in collective decision-making processes. Felsenthal [D. Felsenthal. A Well-Behaved Index of a Priori P-Power for Simple N-Person Games. Homo Oeconomicus, 33, 2016] proposed a power index with a focus on least size winning coalitions, i.e., those coalitions capable of determining the final outcome and with the smallest number of players among all winning coalitions. However, the Felsenthal index overlooks pre-existing affinities between the players, a common and impactful factor in real-world political and economic contexts. This paper introduces the Felsenthal Owen power index, a novel index based on Felsenthal's approach that integrates player affinities using Owen's a priori unions framework. The new index is rigorously characterised by two distinct sets of axiomatic properties. We demonstrate its practical utility by applying it to the International Monetary Fund's voting system, revealing how strategic alliances significantly reshape power distributions. The index thus offers policymakers a more sophisticated tool for measuring influence in complex decision-making scenarios. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:48:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mascareñas-Pazos",
"Alicia",
""
],
[
"Lorenzo-Freire",
"Silvia",
""
],
[
"Alonso-Meijide",
"Jose Maria",
""
]
] |
2507.01622 | Dennis Meier | Ivan N. Ushakov, Mats Topstad, Muhammad Z. Khalid, Niyorjyoti Sharma, Christoph Grams, Ursula Ludacka, Jiali He, Kasper A. Hunnestad, Mohsen Sadeqi-Moqadam, Julia Glaum, Sverre M. Selbach, Joachim Hemberger, Petra Becker, Ladislav Bohat\'y, Amit Kumar, Jorge \'I\~niguez-Gonz\'alez, Antonius T. J. van Helvoort, Dennis Meier | Hybrid antiferroelectric-ferroelectric domain walls in noncollinear antipolar oxides | null | null | null | null | cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | Antiferroelectrics are emerging as advanced functional materials and are fertile ground for unusual electric effects. For example, they enhance the recoverable energy density in energy storage applications and give rise to large electromechanical responses. Here, we demonstrate noncollinearity in dipolar order as an additional degree of freedom, unlocking physical properties that are symmetry-forbidden in classical antiferroelectrics. We show that noncollinear order of electric dipole moments in K$_3$[Nb$_3$O$_6$|(BO$_3$)$_2$] leads to a coexistence of antiferroelectric and ferroelectric behaviors. Besides the double-hysteresis loop observed in antiferroelectrics, a pronounced piezoresponse and electrically switchable domains are observed, separated by atomically sharp and micrometer-long charged domain walls. Hybrid antiferroelectric-ferroelectric responses are expected in a wide range of noncollinear systems, giving a new dimension to the research on antiferroelectrics and multifunctional oxides in general. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:50:39 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ushakov",
"Ivan N.",
""
],
[
"Topstad",
"Mats",
""
],
[
"Khalid",
"Muhammad Z.",
""
],
[
"Sharma",
"Niyorjyoti",
""
],
[
"Grams",
"Christoph",
""
],
[
"Ludacka",
"Ursula",
""
],
[
"He",
"Jiali",
""
],
[
"Hunnestad",
"Kasper A.",
""
],
[
"Sadeqi-Moqadam",
"Mohsen",
""
],
[
"Glaum",
"Julia",
""
],
[
"Selbach",
"Sverre M.",
""
],
[
"Hemberger",
"Joachim",
""
],
[
"Becker",
"Petra",
""
],
[
"Bohatý",
"Ladislav",
""
],
[
"Kumar",
"Amit",
""
],
[
"Íñiguez-González",
"Jorge",
""
],
[
"van Helvoort",
"Antonius T. J.",
""
],
[
"Meier",
"Dennis",
""
]
] |
2507.01624 | Cong Zhou | Cong Zhou and Changsheng You and Shuo Shi and Weidong Mei | Frequency-switching Array Enhanced Physical-Layer Security in Terahertz Bands: A Movable Antenna Perspective | In this paper, we propose to enhance physical-layer security by using a new frequency-switching array, which is equivalent to movable antennas | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a new frequency-switching array (FSA) enhanced physical-layer security (PLS) system in terahertz bands, where the carrier frequency can be flexibly switched and small frequency offsets can be imposed on each antenna at Alice, so as to eliminate information wiretapping by undesired eavesdroppers. First, we analytically show that by flexibly controlling the carrier frequency parameters, FSAs can effectively form uniform/non-uniform sparse arrays, hence resembling movable antennas (MAs) in the control of inter-antenna spacing and providing additional degree-of-freedom (DoF) in the beam control. Although the proposed FSA experiences additional path-gain attenuation in the received signals, it can overcome several hardware and signal processing issues incurred by MAs, such as limited positioning accuracy, considerable response latency, and demanding hardware and energy cost. To shed useful insights, we first consider a secrecy-guaranteed problem with a null-steering constraint for which maximum ratio transmission (MRT) beamformer is considered at Alice and the frequency offsets are set as uniform frequency increment. Interestingly, it is shown that the proposed FSA can flexibly realize null-steering over Eve in both the angular domain (by tuning carrier frequency) and range domain (by controlling per-antenna frequency offset), thereby achieving improved PLS performance. Then, for the general case, we propose an efficient algorithm to solve the formulated non-convex problem by using the block coordinate descent (BCD) and projected gradient ascent (PGA) techniques. Finally, numerical results demonstrate the convergence of the proposed optimization algorithm and its superiority over fixed-position arrays (FPAs) in terms of secrecy-rate performance. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:54:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhou",
"Cong",
""
],
[
"You",
"Changsheng",
""
],
[
"Shi",
"Shuo",
""
],
[
"Mei",
"Weidong",
""
]
] |
2507.01625 | Tomasz Sowi\'nski | Marek Teske and Tomasz Sowi\'nski | Impurity immersed in a two-component few-fermion mixture in a one-dimensional harmonic trap | null | null | null | null | cond-mat.quant-gas | http://creativecommons.org/licenses/by/4.0/ | We investigate a one-dimensional three-component few-fermion mixture confined in a parabolic external trap, where one component contains a single particle acting as an impurity. Focusing on the many-body ground state, we analyze how the interactions between the impurity and the other components influence the system's structure. For fixed interaction strengths within the mixture, we identify a critical interaction strength with the impurity for which the system undergoes a structural transition characterized by substantial change in its spatial features. We explore this transition from the point of view of correlations and ground-state susceptibility. We remarkably find that this transition exhibits unique universality features not previously observed in other systems, highlighting novel many-body properties existing in multi-component fermionic mixtures. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:57:43 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Teske",
"Marek",
""
],
[
"Sowiński",
"Tomasz",
""
]
] |
2507.01626 | Shant Baghram | Zeynab Ashurikisomi, Shant Baghram | Luminosity Function of Galaxies in Voids: A Modification Inspired by Excursion Set Theory | 8 pages, 4 figures. Accepted for publication in the Astrophysical Journal (ApJ) | null | null | null | astro-ph.CO | http://creativecommons.org/licenses/by/4.0/ | In the standard picture of cosmology, the galaxies reside in dark matter (DM) halos. DM halos are distributed in the cosmic web in different environments. The luminosity of the galaxies in different environments can be used as a probe to assess a cosmological model. This study focuses on the properties of galaxies in void regions, where halos typically do not experience extreme conditions. By examining the galaxy luminosity function, we aim to understand the dependence of galaxy properties on their environment and redshift so that later, we could use this as a tool to evaluate cosmological models. We employ the Excursion Set Theory to incorporate parameters related to the number density of DM halos into the luminosity function. Using the Galaxy and Mass Assembly (GAMA) survey and 2dFGRS datasets, we fit our theoretical models to observational data, examining the environmental and redshift dependence of the galaxy luminosity function. Our results indicate that we model the galaxy luminosity function in voids effectively by considering the linear density contrast of the environment and the growth function $D(z)$ for redshift dependence. This study provides a model for the environmental dependence of galaxy luminosity function that offers an improvement in the $\chi ^2$ parameter compared to the previously proposed model in \cite{mcnaught2014galaxy}. Both BIC and AIC tests support the superiority of this model for the void region. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:57:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ashurikisomi",
"Zeynab",
""
],
[
"Baghram",
"Shant",
""
]
] |
2507.01627 | Maeve Hutchinson | Maeve Hutchinson, Radu Jianu, Aidan Slingsby, Jo Wood, Pranava Madhyastha | Chart Question Answering from Real-World Analytical Narratives | This paper has been accepted to the ACL Student Research Workshop (SRW) 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | We present a new dataset for chart question answering (CQA) constructed from visualization notebooks. The dataset features real-world, multi-view charts paired with natural language questions grounded in analytical narratives. Unlike prior benchmarks, our data reflects ecologically valid reasoning workflows. Benchmarking state-of-the-art multimodal large language models reveals a significant performance gap, with GPT-4.1 achieving an accuracy of 69.3%, underscoring the challenges posed by this more authentic CQA setting. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:58:04 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Hutchinson",
"Maeve",
""
],
[
"Jianu",
"Radu",
""
],
[
"Slingsby",
"Aidan",
""
],
[
"Wood",
"Jo",
""
],
[
"Madhyastha",
"Pranava",
""
]
] |
2507.01628 | Zilong He | Zilong He, Pengfei Chen, Hongyu Zhang, Xiaoyun Li, Guangba Yu, Hongyang Chen, Zibin Zheng | DaiFu: In-Situ Crash Recovery for Deep Learning Systems | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Deep learning (DL) systems have been widely adopted in many areas, and are becoming even more popular with the emergence of large language models. However, due to the complex software stacks involved in their development and execution, crashes are unavoidable and common. Crashes severely waste computing resources and hinder development productivity, so efficient crash recovery is crucial. Existing solutions, such as checkpoint-retry, are too heavyweight for fast recovery from crashes caused by minor programming errors or transient runtime errors. Therefore, we present DaiFu, an in-situ recovery framework for DL systems. Through a lightweight code transformation to a given DL system, DaiFu augments it to intercept crashes in situ and enables dynamic and instant updates to its program running context (e.g., code, configurations, and other data) for agile crash recovery. Our evaluation shows that DaiFu helps reduce the restore time for crash recovery, achieving a 1372x speedup compared with state-of-the-art solutions. Meanwhile, the overhead of DaiFu is negligible (under 0.40%). We also construct a benchmark spanning 7 distinct crash scenarios in DL systems, and show the effectiveness of DaiFu in diverse situations. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:58:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"He",
"Zilong",
""
],
[
"Chen",
"Pengfei",
""
],
[
"Zhang",
"Hongyu",
""
],
[
"Li",
"Xiaoyun",
""
],
[
"Yu",
"Guangba",
""
],
[
"Chen",
"Hongyang",
""
],
[
"Zheng",
"Zibin",
""
]
] |
2507.01629 | Tome Eftimov | Tome Eftimov and Peter Koro\v{s}ec | Adaptive Estimation of the Number of Algorithm Runs in Stochastic Optimization | null | null | null | null | cs.NE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Determining the number of algorithm runs is a critical aspect of experimental design, as it directly influences the experiment's duration and the reliability of its outcomes. This paper introduces an empirical approach to estimating the required number of runs per problem instance for accurate estimation of the performance of the continuous single-objective stochastic optimization algorithm. The method leverages probability theory, incorporating a robustness check to identify significant imbalances in the data distribution relative to the mean, and dynamically adjusts the number of runs during execution as an online approach. The proposed methodology was extensively tested across two algorithm portfolios (104 Differential Evolution configurations and the Nevergrad portfolio) and the COCO benchmark suite, totaling 5748000 runs. The results demonstrate 82% - 95% accuracy in estimations across different algorithms, allowing a reduction of approximately 50% in the number of runs without compromising optimization outcomes. This online calculation of required runs not only improves benchmarking efficiency, but also contributes to energy reduction, fostering a more environmentally sustainable computing ecosystem. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:59:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Eftimov",
"Tome",
""
],
[
"Korošec",
"Peter",
""
]
] |
2507.01630 | Yuxiao Wang | Yuxiao Wang, Yu Lei, Zhenao Wei, Weiying Xue, Xinyu Jiang, Nan Zhuang, Qi Liu | Prompt Guidance and Human Proximal Perception for HOT Prediction with Regional Joint Loss | Accepted by ICCV 2025 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The task of Human-Object conTact (HOT) detection involves identifying the specific areas of the human body that are touching objects. Nevertheless, current models are restricted to just one type of image, often leading to too much segmentation in areas with little interaction, and struggling to maintain category consistency within specific regions. To tackle this issue, a HOT framework, termed \textbf{P3HOT}, is proposed, which blends \textbf{P}rompt guidance and human \textbf{P}roximal \textbf{P}erception. To begin with, we utilize a semantic-driven prompt mechanism to direct the network's attention towards the relevant regions based on the correlation between image and text. Then a human proximal perception mechanism is employed to dynamically perceive key depth range around the human, using learnable parameters to effectively eliminate regions where interactions are not expected. Calculating depth resolves the uncertainty of the overlap between humans and objects in a 2D perspective, providing a quasi-3D viewpoint. Moreover, a Regional Joint Loss (RJLoss) has been created as a new loss to inhibit abnormal categories in the same area. A new evaluation metric called ``AD-Acc.'' is introduced to address the shortcomings of existing methods in addressing negative samples. Comprehensive experimental results demonstrate that our approach achieves state-of-the-art performance in four metrics across two benchmark datasets. Specifically, our model achieves an improvement of \textbf{0.7}$\uparrow$, \textbf{2.0}$\uparrow$, \textbf{1.6}$\uparrow$, and \textbf{11.0}$\uparrow$ in SC-Acc., mIoU, wIoU, and AD-Acc. metrics, respectively, on the HOT-Annotated dataset. Code is available at https://github.com/YuxiaoWang-AI/P3HOT. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:59:32 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Wang",
"Yuxiao",
""
],
[
"Lei",
"Yu",
""
],
[
"Wei",
"Zhenao",
""
],
[
"Xue",
"Weiying",
""
],
[
"Jiang",
"Xinyu",
""
],
[
"Zhuang",
"Nan",
""
],
[
"Liu",
"Qi",
""
]
] |
2507.01631 | Camille Billouard | Camille Billouard, Dawa Derksen, Alexandre Constantin, Bruno Vallet | Tile and Slide : A New Framework for Scaling NeRF from Local to Global 3D Earth Observation | Accepted at ICCV 2025 Workshop 3D-VAST (From street to space: 3D Vision Across Altitudes). Version before camera ready. Our code will be made public after the conference | null | null | null | cs.CV cs.AI cs.GR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural Radiance Fields (NeRF) have recently emerged as a paradigm for 3D reconstruction from multiview satellite imagery. However, state-of-the-art NeRF methods are typically constrained to small scenes due to the memory footprint during training, which we study in this paper. Previous work on large-scale NeRFs palliate this by dividing the scene into NeRFs. This paper introduces Snake-NeRF, a framework that scales to large scenes. Our out-of-core method eliminates the need to load all images and networks simultaneously, and operates on a single device. We achieve this by dividing the region of interest into NeRFs that 3D tile without overlap. Importantly, we crop the images with overlap to ensure each NeRFs is trained with all the necessary pixels. We introduce a novel $2\times 2$ 3D tile progression strategy and segmented sampler, which together prevent 3D reconstruction errors along the tile edges. Our experiments conclude that large satellite images can effectively be processed with linear time complexity, on a single GPU, and without compromise in quality. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 11:59:36 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Billouard",
"Camille",
""
],
[
"Derksen",
"Dawa",
""
],
[
"Constantin",
"Alexandre",
""
],
[
"Vallet",
"Bruno",
""
]
] |
2507.01632 | Nils Thorin | Guido Schneider, Nils Thorin | Approximate Peregrine Solitons in Dispersive Nonlinear Wave Equations | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The purpose of this short note is to explain how the existing results on the validity of the NLS approximation can be extended from Sobolev spaces $H^s(\mathbb{R})$ to the spaces of functions $u = v + w$ where $v \in H_{{per}}^s$ and $w \in H^s(\mathbb{R})$. This allows us to use the Peregrine solution of the NLS equation to find freak or rogue wave dynamics in more complicated systems. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:02:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Schneider",
"Guido",
""
],
[
"Thorin",
"Nils",
""
]
] |
2507.01633 | Dmitry Ustalov | Georgii Levtsov and Dmitry Ustalov | Confidence and Stability of Global and Pairwise Scores in NLP Evaluation | 8 pages, accepted at ACL SRW 2025 | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | With the advent of highly capable instruction-tuned neural language models, benchmarking in natural language processing (NLP) is increasingly shifting towards pairwise comparison leaderboards, such as LMSYS Arena, from traditional global pointwise scores (e.g., GLUE, BIG-bench, SWE-bench). This paper empirically investigates the strengths and weaknesses of both global scores and pairwise comparisons to aid decision-making in selecting appropriate model evaluation strategies. Through computational experiments on synthetic and real-world datasets using standard global metrics and the popular Bradley-Terry model for pairwise comparisons, we found that while global scores provide more reliable overall rankings, they can underestimate strong models with rare, significant errors or low confidence. Conversely, pairwise comparisons are particularly effective for identifying strong contenders among models with lower global scores, especially where quality metrics are hard to define (e.g., text generation), though they require more comparisons to converge if ties are frequent. Our code and data are available at https://github.com/HSPyroblast/srw-ranking under a permissive license. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:05:22 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Levtsov",
"Georgii",
""
],
[
"Ustalov",
"Dmitry",
""
]
] |
2507.01634 | Boyuan Sun | Boyuan Sun, Modi Jin, Bowen Yin, Qibin Hou | Depth Anything at Any Condition | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present Depth Anything at Any Condition (DepthAnything-AC), a foundation monocular depth estimation (MDE) model capable of handling diverse environmental conditions. Previous foundation MDE models achieve impressive performance across general scenes but not perform well in complex open-world environments that involve challenging conditions, such as illumination variations, adverse weather, and sensor-induced distortions. To overcome the challenges of data scarcity and the inability of generating high-quality pseudo-labels from corrupted images, we propose an unsupervised consistency regularization finetuning paradigm that requires only a relatively small amount of unlabeled data. Furthermore, we propose the Spatial Distance Constraint to explicitly enforce the model to learn patch-level relative relationships, resulting in clearer semantic boundaries and more accurate details. Experimental results demonstrate the zero-shot capabilities of DepthAnything-AC across diverse benchmarks, including real-world adverse weather benchmarks, synthetic corruption benchmarks, and general benchmarks.
Project Page: https://ghost233lism.github.io/depthanything-AC-page
Code: https://github.com/HVision-NKU/DepthAnythingAC | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:05:57 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Sun",
"Boyuan",
""
],
[
"Jin",
"Modi",
""
],
[
"Yin",
"Bowen",
""
],
[
"Hou",
"Qibin",
""
]
] |
2507.01635 | Chenyu Li | Chenyu Li, Xueping Liang, Xiaorui Gong, Xiu Zhang | EGNInfoLeaker: Unveiling the Risks of Public Key Reuse and User Identity Leakage in Blockchain | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While Ethereum's discovery protocols (Discv4/ Discv5) incorporate robust cryptographic designs to protect user privacy, real-world deployment reveals critical vulnerabilities when users deviate from security guidelines. In this paper, we design a system called EGNInfoLeaker. Our study is the first work that uncovers widespread public key reuse across Ethereum's peer-to-peer networks - a practice that fundamentally undermines the protocol's privacy guarantees. Through systematic analysis of 300 real-world network snapshots, we identify 83 users controlling 483 service nodes via public key reuse, enabling precise de-anonymization through IP correlation. Using evidence collected by EGNInfoLeaker, our Graph-Based Identity Association Algorithm links users to network entities and generates comprehensive user profiles. For User27, it exposes the public key, IP, network ID, location (country/region/city), and ISP/ORG details. The EGNInfoLeaker system demonstrates how such cryptographic misuse transforms theoretical anonymity into practical identity leakage, exposing users to surveillance and targeted attacks. These findings establish that protocol security depends not only on sound design but also on strict user compliance. Going forward, our detection framework provides a foundation for enhancing real-world privacy preservation in decentralized networks. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:07:03 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Li",
"Chenyu",
""
],
[
"Liang",
"Xueping",
""
],
[
"Gong",
"Xiaorui",
""
],
[
"Zhang",
"Xiu",
""
]
] |
2507.01636 | Ghasem Alipoor | Ghasem Alipoor, Karl Skretting | Kernel Recursive Least Squares Dictionary Learning Algorithm | Published in Digital Signal Processing, Volume 141, 2023. DOI: https://doi.org/10.1016/j.dsp.2023.104159 12 pages, 8 figures. Code and data available at: https://github.com/G-Alipoor/kernel-rls-dictionary-learning | Digital Signal Processing, Volume 141, 2023, 104159 | 10.1016/j.dsp.2023.104159 | null | cs.LG eess.SP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We propose an efficient online dictionary learning algorithm for kernel-based sparse representations. In this framework, input signals are nonlinearly mapped to a high-dimensional feature space and represented sparsely using a virtual dictionary. At each step, the dictionary is updated recursively using a novel algorithm based on the recursive least squares (RLS) method. This update mechanism works with single samples or mini-batches and maintains low computational complexity. Experiments on four datasets across different domains show that our method not only outperforms existing online kernel dictionary learning approaches but also achieves classification accuracy close to that of batch-trained models, while remaining significantly more efficient. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:07:35 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Alipoor",
"Ghasem",
""
],
[
"Skretting",
"Karl",
""
]
] |
2507.01637 | Thomas Jan Mikhail | Joachim Kock and Thomas Jan Mikhail | Free loop spaces and the Cauchy--Frobenius Lemma | 4 pages | null | null | CPH-GEOTOP-DNRF151 | math.AT math.CT math.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We upgrade the Cauchy--Frobenius Lemma (`Burnside's Lemma') to a homotopy equivalence of $\infty$-groupoids, essentially given by double counting/Fubini in the free loop space of the quotient. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:08:36 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kock",
"Joachim",
""
],
[
"Mikhail",
"Thomas Jan",
""
]
] |
2507.01638 | Tome Eftimov | Ana Nikolikj and Gabriela Ochoa and Tome Eftimov | Customized Exploration of Landscape Features Driving Multi-Objective Combinatorial Optimization Performance | null | null | null | null | cs.NE cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We present an analysis of landscape features for predicting the performance of multi-objective combinatorial optimization algorithms. We consider features from the recently proposed compressed Pareto Local Optimal Solutions Networks (C-PLOS-net) model of combinatorial landscapes. The benchmark instances are a set of rmnk-landscapes with 2 and 3 objectives and various levels of ruggedness and objective correlation. We consider the performance of three algorithms -- Pareto Local Search (PLS), Global Simple EMO Optimizer (GSEMO), and Non-dominated Sorting Genetic Algorithm (NSGA-II) - using the resolution and hypervolume metrics. Our tailored analysis reveals feature combinations that influence algorithm performance specific to certain landscapes. This study provides deeper insights into feature importance, tailored to specific rmnk-landscapes and algorithms. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:11:41 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Nikolikj",
"Ana",
""
],
[
"Ochoa",
"Gabriela",
""
],
[
"Eftimov",
"Tome",
""
]
] |
2507.01639 | Ilaria Castellano | Ilaria Castellano and Jos\'e Pedro Quintanilha | Geometric invariants of TDLC completions | 38 pages, 2 figures | null | null | null | math.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, Bonn and Sauer showed that, from the point of view of compactness properties, the Schlichting completion of a Hecke pair $(\Gamma,\Lambda)$ behaves precisely as if it were the quotient of $\Gamma$ by $\Lambda$. Motivated by this result, we prove that a similar phenomenon holds for the $\Sigma$-sets. More generally, we relate the $\Sigma$-sets of every TDLC completion of a Hecke pair $(\Gamma,\Lambda)$ to the $\Sigma$-sets of $\Gamma$ whenever $\Lambda$ satisfies suitable compactness properties. We provide applications to TDLC completions of Baumslag-Solitar groups and certain groups of upper triangular matrices studied by Schesler. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:11:46 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Castellano",
"Ilaria",
""
],
[
"Quintanilha",
"José Pedro",
""
]
] |
2507.01640 | Anath Bandhu Das | Anath Bandhu Das, Sangita Dutta, and Pinaki Pal | Effect of phase-lag on synchronization in adaptive multilayer networks with higher-order interactions | 14 pages, 14 figures | null | null | null | nlin.AO math.DS nlin.CD physics.soc-ph | http://creativecommons.org/licenses/by/4.0/ | We investigate the transition to synchronization in adaptive multilayer networks with higher-order interactions both analytically and numerically in the presence of phase frustration ($\beta$). The higher order topology consists of pairwise and triadic couplings. The analytical framework for the investigation is based on the Ott-Antonsen ansatz which leads to a convenient low-dimensional model. Extensive bifurcation analysis of the low-dimensional model and the numerical simulation of the full networks are performed to explore the paths to synchronization. The combined analysis shows a complex dependence of the transition to synchronization on adaptation exponents, coupling strengths, phase lag parameter, and multilayer configuration. Various types of transitions to synchronization, namely continuous, tiered, and explosive, are exhibited by the system in different regions of the parameter space. In all the cases, a satisfactory match between the low-dimensional model and the numerical simulation results is observed. The origin of different transitions to synchronization is clearly understood using the low-dimensional model. Exploration of a wide region of the parameter space suggests that the phase frustration parameter inhibits tired as well as explosive synchronization transitions for fixed triadic coupling strength ($K_2$). On the other hand, discontinuous transition is promoted by the phase frustration parameter for fixed pairwise coupling strength ($K_1$). Moreover, the exponent of the adaptation function with the pairwise coupling decreases the width of the hysteresis, despite the dominance of the higher-order coupling for fixed $\beta$ and $K_2$. While, the exponent of the function adapted with higher-order coupling shows the opposite effect, it promotes bistability in spite of dominance of pairwise coupling strength for fixed $\beta$, and $K_1$. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:11:54 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Das",
"Anath Bandhu",
""
],
[
"Dutta",
"Sangita",
""
],
[
"Pal",
"Pinaki",
""
]
] |
2507.01642 | Jens Schr\"oder | Jens Schr\"oder and Emil Wiedemann | On the Vanishing Viscosity Limit for Inhomogeneous Incompressible Navier-Stokes Equations on Bounded Domains | 14 pages | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we study the vanishing viscosity limit for the inhomogeneous incompressible Navier-Stokes equations on bounded domains with no-slip boundary condition in two or three space dimensions. We show that, under suitable assumptions on the density, we can establish the convergence in energy space of Leray-Hopf type solutions of the Navier-Stokes equation to a smooth solution of the Euler equations if and only if the energy dissipation vanishes on a boundary layer with thickness proportional to the viscosity. This extends Kato's criterion for homogeneous Navier-Stokes equations to the inhomogeneous case. We use a new relative energy functional in our proof. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:16:49 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Schröder",
"Jens",
""
],
[
"Wiedemann",
"Emil",
""
]
] |
2507.01643 | Dingkang Yang | Weijie Yin, Dingkang Yang, Hongyuan Dong, Zijian Kang, Jiacong Wang, Xiao Liang, Chao Feng, Jiao Ran | SAILViT: Towards Robust and Generalizable Visual Backbones for MLLMs via Gradual Feature Refinement | We release SAILViT, a series of versatile vision foundation models | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision Transformers (ViTs) are essential as foundation backbones in establishing the visual comprehension capabilities of Multimodal Large Language Models (MLLMs). Although most ViTs achieve impressive performance through image-text pair-based contrastive learning or self-supervised mechanisms, they struggle to engage in connector-based co-training directly with LLMs due to potential parameter initialization conflicts and modality semantic gaps. To address the above challenges, this paper proposes SAILViT, a gradual feature learning-enhanced ViT for facilitating MLLMs to break through performance bottlenecks in complex multimodal interactions. SAILViT achieves coarse-to-fine-grained feature alignment and world knowledge infusion with gradual feature refinement, which better serves target training demands. We perform thorough empirical analyses to confirm the powerful robustness and generalizability of SAILViT across different dimensions, including parameter sizes, model architectures, training strategies, and data scales. Equipped with SAILViT, existing MLLMs show significant and consistent performance improvements on the OpenCompass benchmark across extensive downstream tasks. SAILViT series models are released at https://huggingface.co/BytedanceDouyinContent. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:17:23 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Yin",
"Weijie",
""
],
[
"Yang",
"Dingkang",
""
],
[
"Dong",
"Hongyuan",
""
],
[
"Kang",
"Zijian",
""
],
[
"Wang",
"Jiacong",
""
],
[
"Liang",
"Xiao",
""
],
[
"Feng",
"Chao",
""
],
[
"Ran",
"Jiao",
""
]
] |
2507.01644 | Miguel O'Malley | Miguel O'Malley | Dance Dance ConvLSTM | 15 pages, 9 figures, 4 tables | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | \textit{Dance Dance Revolution} is a rhythm game consisting of songs and accompanying choreography, referred to as charts. Players press arrows on a device referred to as a dance pad in time with steps determined by the song's chart. In 2017, the authors of Dance Dance Convolution (DDC) developed an algorithm for the automatic generation of \textit{Dance Dance Revolution} charts, utilizing a CNN-LSTM architecture. We introduce Dance Dance ConvLSTM (DDCL), a new method for the automatic generation of DDR charts using a ConvLSTM based model, which improves upon the DDC methodology and substantially increases the accuracy of chart generation. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:17:33 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"O'Malley",
"Miguel",
""
]
] |
2507.01645 | Rifki Afina Putri | Rifki Afina Putri | Adapting Language Models to Indonesian Local Languages: An Empirical Study of Language Transferability on Zero-Shot Settings | AMLDS 2025 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we investigate the transferability of pre-trained language models to low-resource Indonesian local languages through the task of sentiment analysis. We evaluate both zero-shot performance and adapter-based transfer on ten local languages using models of different types: a monolingual Indonesian BERT, multilingual models such as mBERT and XLM-R, and a modular adapter-based approach called MAD-X. To better understand model behavior, we group the target languages into three categories: seen (included during pre-training), partially seen (not included but linguistically related to seen languages), and unseen (absent and unrelated in pre-training data). Our results reveal clear performance disparities across these groups: multilingual models perform best on seen languages, moderately on partially seen ones, and poorly on unseen languages. We find that MAD-X significantly improves performance, especially for seen and partially seen languages, without requiring labeled data in the target language. Additionally, we conduct a further analysis on tokenization and show that while subword fragmentation and vocabulary overlap with Indonesian correlate weakly with prediction quality, they do not fully explain the observed performance. Instead, the most consistent predictor of transfer success is the model's prior exposure to the language, either directly or through a related language. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:17:55 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Putri",
"Rifki Afina",
""
]
] |
2507.01646 | Randy Dobler | Randy Dobler, Juliana O. Costa, Marcelo D. Alloy, D\'ebora P. Menezes | Bayesian Analysis of Non-extensive Parameters in Au-Au Collisions | null | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, a Bayesian statistical framework is employed to analyze particle yield ratios in Au-Au collisions, utilizing Non-Extensive Statistics (NES). Through Markov Chain Monte Carlo (MCMC) sampling, we systematically estimate key parameters, including the non-extensive factor $q$, temperature $T$, and chemical potential $\mu$. Our analysis confirms previous findings highlighting the suitability and robustness of Bayesian methods in describing heavy-ion collision data. A subsequent Bayes factor analysis does not provide definitive evidence favoring an Excluded Volume Hadron Resonance Gas (EV-HRG) model over the simpler NES approach. Overall, these results suggest that combining NES with Bayesian inference can effectively model particle distributions and improve parameter estimation accuracy, demonstrating the potential of this approach for future studies on relativistic nuclear interactions. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:20:24 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Dobler",
"Randy",
""
],
[
"Costa",
"Juliana O.",
""
],
[
"Alloy",
"Marcelo D.",
""
],
[
"Menezes",
"Débora P.",
""
]
] |
2507.01647 | Niclas Bernhoff | Niclas Bernhoff, Stephane Brull, Eddie Wadbro | The half-space problem of evaporation and condensation for polyatomic gases and entropy inequalities | 30 pages, 7 figures | null | null | null | math.AP math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study investigates the steady Boltzmann equation in one spatial variable for a polyatomic single-component gas in a half-space. Inflow boundary conditions are assumed at the half-space boundary, where particles entering the half-space are distributed as a Maxwellian, an equilibrium distribution characterized by macroscopic parameters of the boundary. At the far end, the gas tends to an equilibrium distribution, which is also Maxwellian. Using conservation laws and an entropy inequality, we derive relations between the macroscopic parameters of the boundary and at infinity required for the existence of solutions. The relations vary depending on the sign of the Mach number at infinity, which dictates whether evaporation or condensation takes place at the interface between the gas and the condensed phase. We explore the obtained relations numerically. This investigation reveals that, although the relations are qualitatively similar for various internal degrees of freedom of the gas, clear quantitative differences exist. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:20:26 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bernhoff",
"Niclas",
""
],
[
"Brull",
"Stephane",
""
],
[
"Wadbro",
"Eddie",
""
]
] |
2507.01648 | Petros Laccotripes | Petros Laccotripes, Junyang Huang, Ginny Shooter, Andrea Barbiero, Matthew S. Winnel, David A. Ritchie, Andrew J. Shields, Tina Muller, R. Mark Stevenson | An entangled photon source for the telecom C-band based on a semiconductor-confined spin | 18 pages, 4 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multiphoton entangled states are a key resource for quantum networks and measurement-based quantum computation. Scalable protocols for generating such states using solid-state spin-photon interfaces have recently emerged, but practical implementations have so far relied on emitters operating at short wavelengths, incompatible with low-loss fibre transmission. Here, we take a key step towards the generation of telecom wavelength multi-qubit entangled states using an InAs/InP quantum dot. After establishing that all essential criteria for generating cluster states using a ground state spin as the entangler are satisfied, we implement a scalable protocol to entangle the resident spin with sequentially emitted photons directly in the telecom C-band. We demonstrate a two-qubit (spin-photon) entanglement fidelity of $59.5\pm 8.7\%$ and a lower bound of three-qubit (spin-photon-photon) entanglement fidelity of $52.7\pm 11.4\%$. Our results close the performance gap between short-wavelength quantum dot systems and the existing telecom infrastructure, establishing a route towards practical large photonic cluster states for fibre-based quantum network applications. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:22:19 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Laccotripes",
"Petros",
""
],
[
"Huang",
"Junyang",
""
],
[
"Shooter",
"Ginny",
""
],
[
"Barbiero",
"Andrea",
""
],
[
"Winnel",
"Matthew S.",
""
],
[
"Ritchie",
"David A.",
""
],
[
"Shields",
"Andrew J.",
""
],
[
"Muller",
"Tina",
""
],
[
"Stevenson",
"R. Mark",
""
]
] |
2507.01649 | Yoav Gelberg | Yoav Gelberg, Yam Eitan, Aviv Navon, Aviv Shamsian, Theo (Moe) Putterman, Michael Bronstein, Haggai Maron | GradMetaNet: An Equivariant Architecture for Learning on Gradients | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Gradients of neural networks encode valuable information for optimization, editing, and analysis of models. Therefore, practitioners often treat gradients as inputs to task-specific algorithms, e.g. for pruning or optimization. Recent works explore learning algorithms that operate directly on gradients but use architectures that are not specifically designed for gradient processing, limiting their applicability. In this paper, we present a principled approach for designing architectures that process gradients. Our approach is guided by three principles: (1) equivariant design that preserves neuron permutation symmetries, (2) processing sets of gradients across multiple data points to capture curvature information, and (3) efficient gradient representation through rank-1 decomposition. Based on these principles, we introduce GradMetaNet, a novel architecture for learning on gradients, constructed from simple equivariant blocks. We prove universality results for GradMetaNet, and show that previous approaches cannot approximate natural gradient-based functions that GradMetaNet can. We then demonstrate GradMetaNet's effectiveness on a diverse set of gradient-based tasks on MLPs and transformers, such as learned optimization, INR editing, and estimating loss landscape curvature. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:22:39 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gelberg",
"Yoav",
"",
"Moe"
],
[
"Eitan",
"Yam",
"",
"Moe"
],
[
"Navon",
"Aviv",
"",
"Moe"
],
[
"Shamsian",
"Aviv",
"",
"Moe"
],
[
"Theo",
"",
"",
"Moe"
],
[
"Putterman",
"",
""
],
[
"Bronstein",
"Michael",
""
],
[
"Maron",
"Haggai",
""
]
] |
2507.01651 | Sylvain Fontaine | Sylvain Fontaine | A Dynamical Cartography of the Epistemic Diffusion of Artificial Intelligence in Neuroscience | null | null | null | null | cs.DL physics.soc-ph q-bio.NC | http://creativecommons.org/licenses/by/4.0/ | Neuroscience and AI have an intertwined history, largely relayed in the literature of both fields. In recent years, due to the engineering orientations of AI research and the monopoly of industry for its large-scale applications, the mutual expansion of neuroscience and AI in fundamental research seems challenged. In this paper, we bring some empirical evidences that, on the contrary, AI and neuroscience are continuing to grow together, but with a pronounced interest in the fields of study related to neurodegenerative diseases since the 1990s. With a temporal knowledge cartography of neuroscience drawn with advanced document embedding techniques, we draw the dynamical shaping of the discipline since the 1970s and identified the conceptual articulation of AI with this particular subfield mentioned before. However, a further analysis of the underlying citation network of the studied corpus shows that the produced AI technologies remain confined in the different subfields and are not transferred from one subfield to another. This invites us to discuss the genericity capability of AI in the context of an intradisciplinary development, especially in the diffusion of its associated metrology. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:24:44 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Fontaine",
"Sylvain",
""
]
] |
2507.01652 | Yuxin Mao | Yuxin Mao, Zhen Qin, Jinxing Zhou, Hui Deng, Xuyang Shen, Bin Fan, Jing Zhang, Yiran Zhong, Yuchao Dai | Autoregressive Image Generation with Linear Complexity: A Spatial-Aware Decay Perspective | null | null | null | null | cs.CV cs.AI cs.MM | http://creativecommons.org/licenses/by/4.0/ | Autoregressive (AR) models have garnered significant attention in image generation for their ability to effectively capture both local and global structures within visual data. However, prevalent AR models predominantly rely on the transformer architectures, which are beset by quadratic computational complexity concerning input sequence length and substantial memory overhead due to the necessity of maintaining key-value caches. Although linear attention mechanisms have successfully reduced this burden in language models, our initial experiments reveal that they significantly degrade image generation quality because of their inability to capture critical long-range dependencies in visual data. We propose Linear Attention with Spatial-Aware Decay (LASAD), a novel attention mechanism that explicitly preserves genuine 2D spatial relationships within the flattened image sequences by computing position-dependent decay factors based on true 2D spatial location rather than 1D sequence positions. Based on this mechanism, we present LASADGen, an autoregressive image generator that enables selective attention to relevant spatial contexts with linear complexity. Experiments on ImageNet show LASADGen achieves state-of-the-art image generation performance and computational efficiency, bridging the gap between linear attention's efficiency and spatial understanding needed for high-quality generation. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:27:06 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mao",
"Yuxin",
""
],
[
"Qin",
"Zhen",
""
],
[
"Zhou",
"Jinxing",
""
],
[
"Deng",
"Hui",
""
],
[
"Shen",
"Xuyang",
""
],
[
"Fan",
"Bin",
""
],
[
"Zhang",
"Jing",
""
],
[
"Zhong",
"Yiran",
""
],
[
"Dai",
"Yuchao",
""
]
] |
2507.01653 | Yuran Wang | Yuran Wang, Yingping Liang, Yutao Hu, Ying Fu | RobuSTereo: Robust Zero-Shot Stereo Matching under Adverse Weather | accepted by ICCV25 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning-based stereo matching models struggle in adverse weather conditions due to the scarcity of corresponding training data and the challenges in extracting discriminative features from degraded images. These limitations significantly hinder zero-shot generalization to out-of-distribution weather conditions. In this paper, we propose \textbf{RobuSTereo}, a novel framework that enhances the zero-shot generalization of stereo matching models under adverse weather by addressing both data scarcity and feature extraction challenges. First, we introduce a diffusion-based simulation pipeline with a stereo consistency module, which generates high-quality stereo data tailored for adverse conditions. By training stereo matching models on our synthetic datasets, we reduce the domain gap between clean and degraded images, significantly improving the models' robustness to unseen weather conditions. The stereo consistency module ensures structural alignment across synthesized image pairs, preserving geometric integrity and enhancing depth estimation accuracy. Second, we design a robust feature encoder that combines a specialized ConvNet with a denoising transformer to extract stable and reliable features from degraded images. The ConvNet captures fine-grained local structures, while the denoising transformer refines global representations, effectively mitigating the impact of noise, low visibility, and weather-induced distortions. This enables more accurate disparity estimation even under challenging visual conditions. Extensive experiments demonstrate that \textbf{RobuSTereo} significantly improves the robustness and generalization of stereo matching models across diverse adverse weather scenarios. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:27:53 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Wang",
"Yuran",
""
],
[
"Liang",
"Yingping",
""
],
[
"Hu",
"Yutao",
""
],
[
"Fu",
"Ying",
""
]
] |
2507.01654 | Ad\'in Ram\'irez Rivera | Martine Hjelkrem-Tan and Marius Aasan and Gabriel Y. Arteaga and Ad\'in Ram\'irez Rivera | SPoT: Subpixel Placement of Tokens in Vision Transformers | To appear in Workshop on Efficient Computing under Limited Resources: Visual Computing (ICCV 2025). Code available at https://github.com/dsb-ifi/SPoT | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Vision Transformers naturally accommodate sparsity, yet standard tokenization methods confine features to discrete patch grids. This constraint prevents models from fully exploiting sparse regimes, forcing awkward compromises. We propose Subpixel Placement of Tokens (SPoT), a novel tokenization strategy that positions tokens continuously within images, effectively sidestepping grid-based limitations. With our proposed oracle-guided search, we uncover substantial performance gains achievable with ideal subpixel token positioning, drastically reducing the number of tokens necessary for accurate predictions during inference. SPoT provides a new direction for flexible, efficient, and interpretable ViT architectures, redefining sparsity as a strategic advantage rather than an imposed limitation. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:30:32 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Hjelkrem-Tan",
"Martine",
""
],
[
"Aasan",
"Marius",
""
],
[
"Arteaga",
"Gabriel Y.",
""
],
[
"Rivera",
"Adín Ramírez",
""
]
] |
2507.01655 | Stefan Ivanov | Stefan Ivanov and Luis Ugarte | Almost Calabi-Yau with torsion 6-manifolds and the instanton condition | 22 pages, no figures | null | null | null | math.DG hep-th | http://creativecommons.org/licenses/by/4.0/ | It is observed that on a compact almost complex Calabi-Yau with torsion (ACYT) 6-manifold with co-closed Lee form the curvature of the torsion connection is an $SU(3)$-instanton if and only if the torsion is parallel with respect to the torsion connection. The same conclusion holds for any (non necessarily compact) balanced ACYT 6-manifold. In particular, on a CYT 6-manifold the Strominger-Bismut connection is an $SU(3)$-instanton if and only if the torsion is parallel with respect to the Strominger-Bismut connection provided either the CYT 6-manifold is compact with co-closed Lee form or it is a balanced CYT 6-manifold. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:32:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ivanov",
"Stefan",
""
],
[
"Ugarte",
"Luis",
""
]
] |
2507.01656 | Finn Bartsch | Finn Bartsch | The Kobayashi pseudometric in the presence of log-terminal singularities | 3 pages, comments welcome! | null | null | null | math.AG math.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We show that the Kobayashi pseudometric is well-behaved under resolution of log-terminal singularities. This answers a question of Kamenova and Lehn. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:33:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bartsch",
"Finn",
""
]
] |
2507.01657 | Vivien Ducros | Vivien Ducros and Claudio Zandron | Symport/Antiport P Systems with Membrane Separation Characterize P^(#P) | null | null | null | null | cs.CC | http://creativecommons.org/licenses/by/4.0/ | Membrane systems represent a computational model that operates in a distributed and parallel manner, inspired by the behavior of biological cells. These systems feature objects that transform within a nested membrane structure. This research concentrates on a specific type of these systems, based on cellular symport/antiport communication of chemicals.
Results in the literature show that systems of this type that also allow cell division can solve PSPACE problems. In our study, we investigate systems that use membrane separation instead of cell division, for which only limited results are available. Notably, it has been shown that any problem solvable by such systems in polynomial time falls within the complexity class P^(#P).
By implementing a system solving MIDSAT, a P^(#P)-complete problem, we demonstrate that the reverse inclusion is true as well, thus providing an exact characterization of the problem class solvable by P systems with symport/antiport and membrane separation.
Moreover, our implementation uses rules of length at most three. With this limit, systems were known to be able to solve NP-complete problems, whereas limiting the rules by length two, they characterize P. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:33:27 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ducros",
"Vivien",
""
],
[
"Zandron",
"Claudio",
""
]
] |
2507.01658 | Jacek Sekutowicz | J. Sekutowicz, D. Bazyl, E. Vogel, D. Reschke, A. Brinkmann, D. Klinke, D. Kostin, T. Ramm, A. Sulimov, H. Weise, M. Wiencek | Cathodes and Shape Modification of Cavity for DESY Superconducting Photoinjector | 11 pages, 14 figures | null | null | null | physics.acc-ph | http://creativecommons.org/licenses/by/4.0/ | Four DESY prototypes of the L-band superconducting RF photoinjector cavity demonstrated on-axis peak gradients above 55MV/m during multiple vertical cryogenic tests. Two of these prototypes,16G09 and 16G10, achieved these gradients with both superconducting and normal-conducting metallic cathodes, fabricated from either high-purity niobium or lower-purity copper. The DESY photoinjector, under development for over two decades as a continuous-wave electron source for FELs, differs from other SRF injectors in that its metallic cathode plug is attached directly to the cavity backplate, exposing the emitting surface to the high electric field within the cavity. This design obviates the need for a choke filter or load-lock system. The initial 1.6-cell cavity geometry was derived from the Low-Loss design developed for the CEBAF 12GeV upgrade, scaled from 1.5GHz to 1.3GHz. This shape was later replaced with the current High Gradient TESLA profile. In this report, we discuss current cathode options and present modifications to the cavity shape aimed at significantly reducing the electric field near the cathode opening | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:40:47 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Sekutowicz",
"J.",
""
],
[
"Bazyl",
"D.",
""
],
[
"Vogel",
"E.",
""
],
[
"Reschke",
"D.",
""
],
[
"Brinkmann",
"A.",
""
],
[
"Klinke",
"D.",
""
],
[
"Kostin",
"D.",
""
],
[
"Ramm",
"T.",
""
],
[
"Sulimov",
"A.",
""
],
[
"Weise",
"H.",
""
],
[
"Wiencek",
"M.",
""
]
] |
2507.01659 | Wilke van der Schee | Govert Nijs, Wilke van der Schee | Transmutation of $^{16}$O and $^{20}$Ne at the Large Hadron Collider | 3 pages, 3 figures and an appendix. The Trajectum code can be found at https://sites.google.com/view/govertnijs/trajectum | null | null | CERN-TH-2025-127 | nucl-th nucl-ex | http://creativecommons.org/licenses/by/4.0/ | In July 2025 the Large Hadron Collider (LHC) will collide $^{16}$O$^{16}$O and $^{20}$Ne$^{20}$Ne isotopes in a quest to understand the physics of ultrarelativistic light ion collisions. One particular feature is that there are many smaller isotopes with the exact same charge over mass ratio that potentially can be produced and contaminate the beam composition. Using the Trajectum framework together with the GEMINI code we provide an estimate of the production cross-section and its consequences. A potential benefit could be the interesting measurement of the multiplicity and mean transverse momentum of $^{16}$O$^{4}$He collisions. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:41:08 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Nijs",
"Govert",
""
],
[
"van der Schee",
"Wilke",
""
]
] |
2507.01660 | Yilin Zou | Yilin Zou and Fanghua Jiang | Re-examining the Legendre-Gauss-Lobatto Pseudospectral Methods for Optimal Control | null | null | null | null | eess.SY cs.SY math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pseudospectral methods represent an efficient approach for solving optimal control problems. While Legendre-Gauss-Lobatto (LGL) collocation points have traditionally been considered inferior to Legendre-Gauss (LG) and Legendre-Gauss-Radau (LGR) points in terms of convergence properties, this paper presents a rigorous re-examination of LGL-based methods. We introduce an augmented formulation that enhances the standard LGL collocation approach by incorporating an additional degree of freedom (DOF) into the interpolation structure. We demonstrate that this augmented formulation is mathematically equivalent to the integral formulation of the LGL collocation method. Through analytical derivation, we establish that the adjoint system in both the augmented differential and integral formulations corresponds to a Lobatto IIIB discontinuous collocation method for the costate vector, thereby resolving the previously reported convergence issues. Our comparative analysis of LG, LGR, and LGL collocation methods reveals significant advantages of the improved LGL approach in terms of discretized problem dimensionality and symplectic integration properties. Numerical examples validate our theoretical findings, demonstrating that the proposed LGL-based method achieves comparable accuracy to LG and LGR methods while offering superior computational performance for long-horizon optimal control problems due to the preservation of symplecticity. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:41:33 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zou",
"Yilin",
""
],
[
"Jiang",
"Fanghua",
""
]
] |
2507.01661 | Lav Kumar Singh | Lav Kumar Singh and Aljosa Peperko | On some Fr\'echet spaces associated to the functions satisfying Mulholland inequality | null | null | null | null | math.FA math.MG | http://creativecommons.org/licenses/by/4.0/ | In this article we explore a new growth condition on Young functions, which we call Mulholland condition, pertaining to the mathematician H.P Mulholland, who studied these functions for the first time, albeit in a different context. We construct a non-trivial Young function $\Omega$ which satisfies Mulholland condition and $\Delta_2$-condition. We then associate exotic $F$-norms to the vector space $X_1\oplus X_2$, where $X_1$ and $X_2$ are Banach spaces, using the function $\Omega$. This $F$-spaces contains the Banach space $X_1$ and $X_2$ as a maximal Banach subspace. Further, the Banach envelope $(X_1\oplus X_2,||.||_{\Omega_o})$ of this $F$-space corresponds to the Young function $\Omega_o$ who characteristic function is an asymptotic line to the characteristic function of the Young function $\Omega$. Thus these $F$-spaces serves as "interpolation space" for Banach spaces $X_1$ and $(X_1\oplus X_2, ||.||_{\Omega_o})$ in some sense. These $F$-space are well behaved in regards to Hahn-Banach extension property, which is lacking in classical $F$-spaces like $L^p$ and $H^p$ for $0<p<1$. Towards the end, some direct sums for Orlicz spaces are discussed. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:42:15 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Singh",
"Lav Kumar",
""
],
[
"Peperko",
"Aljosa",
""
]
] |
2507.01662 | Alan Dow | Alan Dow | Not OCA and products of Frechet spaces | 16 pages | null | null | null | math.GN math.LO | http://creativecommons.org/licenses/by/4.0/ | We continue the investigation of the question of whether the product of two countable Fr\'echet spaces must be M-separable. We are especially interested in this question in the presence of Martin's Axiom. The question has been shown to be independent of Martin's Axiom but only in models in which $\mathfrak c\leq\omega_2$. In fact, OCA implies an affirmative answer. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:44:57 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Dow",
"Alan",
""
]
] |
2507.01663 | Zhenyu Han | Zhenyu Han, Ansheng You, Haibo Wang, Kui Luo, Guang Yang, Wenqi Shi, Menglong Chen, Sicheng Zhang, Zeshun Lan, Chunshi Deng, Huazhong Ji, Wenjie Liu, Yu Huang, Yixiang Zhang, Chenyi Pan, Jing Wang, Xin Huang, Chunsheng Li, Jianping Wu | AsyncFlow: An Asynchronous Streaming RL Framework for Efficient LLM Post-Training | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Reinforcement learning (RL) has become a pivotal technology in the post-training phase of large language models (LLMs). Traditional task-colocated RL frameworks suffer from significant scalability bottlenecks, while task-separated RL frameworks face challenges in complex dataflows and the corresponding resource idling and workload imbalance. Moreover, most existing frameworks are tightly coupled with LLM training or inference engines, making it difficult to support custom-designed engines. To address these challenges, we propose AsyncFlow, an asynchronous streaming RL framework for efficient post-training. Specifically, we introduce a distributed data storage and transfer module that provides a unified data management and fine-grained scheduling capability in a fully streamed manner. This architecture inherently facilitates automated pipeline overlapping among RL tasks and dynamic load balancing. Moreover, we propose a producer-consumer-based asynchronous workflow engineered to minimize computational idleness by strategically deferring parameter update process within staleness thresholds. Finally, the core capability of AsynFlow is architecturally decoupled from underlying training and inference engines and encapsulated by service-oriented user interfaces, offering a modular and customizable user experience. Extensive experiments demonstrate an average of 1.59 throughput improvement compared with state-of-the-art baseline. The presented architecture in this work provides actionable insights for next-generation RL training system designs. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:45:34 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Han",
"Zhenyu",
""
],
[
"You",
"Ansheng",
""
],
[
"Wang",
"Haibo",
""
],
[
"Luo",
"Kui",
""
],
[
"Yang",
"Guang",
""
],
[
"Shi",
"Wenqi",
""
],
[
"Chen",
"Menglong",
""
],
[
"Zhang",
"Sicheng",
""
],
[
"Lan",
"Zeshun",
""
],
[
"Deng",
"Chunshi",
""
],
[
"Ji",
"Huazhong",
""
],
[
"Liu",
"Wenjie",
""
],
[
"Huang",
"Yu",
""
],
[
"Zhang",
"Yixiang",
""
],
[
"Pan",
"Chenyi",
""
],
[
"Wang",
"Jing",
""
],
[
"Huang",
"Xin",
""
],
[
"Li",
"Chunsheng",
""
],
[
"Wu",
"Jianping",
""
]
] |
2507.01664 | Hector Gramaglia | Hector Gramaglia | Globality and Regions | null | null | null | null | cs.PL | http://creativecommons.org/licenses/by/4.0/ | We obtain a characterization of global variables by unifying abstraction with region abstraction in a region-based language. More precisely, in a previous work a language called global was presented, whose virtue is to provide a conceptually clear way of introducing imperative operations in a functional language. Memory safety is provided by the concept of linear protection, which connects the global system to a linear one. In this paper we show that the concept of global variable provided by the global language arises from the Tofte and Talping's region language through the unification of abstraction and region abstraction. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:45:50 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gramaglia",
"Hector",
""
]
] |
2507.01665 | Kazuhiro Hikami | Kazuhiro Hikami | Askey-Wilson polynomial and genus-two skein module | 10 pages, 1 figure | null | null | null | math.QA math-ph math.GT math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Based on the q-difference operators for the genus-two skein algebra, we show the correspondence between the reduced Askey-Wilson polynomials and the skein module of the genus-two handlebody. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:47:59 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Hikami",
"Kazuhiro",
""
]
] |
2507.01666 | Andreas Michels | I. Titov, M. Bersweiler, M.P. Adams, E.P. Sinaga, V. Rai, S. Liscak, M. Lahr, T.L. Schmidt, V.M. Kuchkin, A. Haller, K. Suzuki, N.-J. Steinke, D. Alba Venero, D. Honecker, J. Kohlbrecher, L.F. Barquin, and A. Michels | Spin-disorder-induced angular anisotropy in polarized magnetic neutron scattering | null | null | null | null | cond-mat.mes-hall cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | We experimentally report a hitherto unseen angular anisotropy in the polarized small-angle neutron scattering (SANS) cross section of a magnetically strongly inhomogeneous material. Based on an analytical prediction using micromagnetic theory, the difference between the spin-up and spin-down SANS cross sections is expected to show a spin-disorder-induced anisotropy. The effect is particularly pronounced in inhomogeneous magnetic materials, such as nanoporous ferromagnets, magnetic nanocomposites, or steels, which exhibit large nanoscale jumps in the saturation magnetization at internal pore-matrix or particle-matrix interfaces. Analysis of the experimental neutron data constitutes a method for determining the exchange-stiffness constant. Our results are generic to the nuclear-magnetic interference terms contained in the polarized magnetic neutron scattering cross section and might also be of relevance to other neutron techniques. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:48:01 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Titov",
"I.",
""
],
[
"Bersweiler",
"M.",
""
],
[
"Adams",
"M. P.",
""
],
[
"Sinaga",
"E. P.",
""
],
[
"Rai",
"V.",
""
],
[
"Liscak",
"S.",
""
],
[
"Lahr",
"M.",
""
],
[
"Schmidt",
"T. L.",
""
],
[
"Kuchkin",
"V. M.",
""
],
[
"Haller",
"A.",
""
],
[
"Suzuki",
"K.",
""
],
[
"Steinke",
"N. -J.",
""
],
[
"Venero",
"D. Alba",
""
],
[
"Honecker",
"D.",
""
],
[
"Kohlbrecher",
"J.",
""
],
[
"Barquin",
"L. F.",
""
],
[
"Michels",
"A.",
""
]
] |
2507.01667 | Gianluca Monaci | Gianluca Monaci, Philippe Weinzaepfel, Christian Wolf | What does really matter in image goal navigation? | null | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image goal navigation requires two different skills: firstly, core navigation skills, including the detection of free space and obstacles, and taking decisions based on an internal representation; and secondly, computing directional information by comparing visual observations to the goal image. Current state-of-the-art methods either rely on dedicated image-matching, or pre-training of computer vision modules on relative pose estimation. In this paper, we study whether this task can be efficiently solved with end-to-end training of full agents with RL, as has been claimed by recent work. A positive answer would have impact beyond Embodied AI and allow training of relative pose estimation from reward for navigation alone. In a large study we investigate the effect of architectural choices like late fusion, channel stacking, space-to-depth projections and cross-attention, and their role in the emergence of relative pose estimators from navigation training. We show that the success of recent methods is influenced up to a certain extent by simulator settings, leading to shortcuts in simulation. However, we also show that these capabilities can be transferred to more realistic setting, up to some extend. We also find evidence for correlations between navigation performance and probed (emerging) relative pose estimation performance, an important sub skill. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:50:26 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Monaci",
"Gianluca",
""
],
[
"Weinzaepfel",
"Philippe",
""
],
[
"Wolf",
"Christian",
""
]
] |
2507.01668 | Tome Eftimov | Gjorgjina Cenikj and Ga\v{s}per Petelin and Tome Eftimov | Comparing Optimization Algorithms Through the Lens of Search Behavior Analysis | null | null | null | null | cs.NE cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The field of numerical optimization has recently seen a surge in the development of "novel" metaheuristic algorithms, inspired by metaphors derived from natural or human-made processes, which have been widely criticized for obscuring meaningful innovations and failing to distinguish themselves from existing approaches. Aiming to address these concerns, we investigate the applicability of statistical tests for comparing algorithms based on their search behavior. We utilize the cross-match statistical test to compare multivariate distributions and assess the solutions produced by 114 algorithms from the MEALPY library. These findings are incorporated into an empirical analysis aiming to identify algorithms with similar search behaviors. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:51:27 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Cenikj",
"Gjorgjina",
""
],
[
"Petelin",
"Gašper",
""
],
[
"Eftimov",
"Tome",
""
]
] |
2507.01669 | Matthias Franz | Matthias Franz | The Szczarba map and the cubical cobar construction | 26 pages | null | null | null | math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider a twisting function from a 1-reduced simplicial set $X$ to a simplicial group $G$. We prove in detail that the associated Szczarba operators induce a simplicial map from the triangulation of the cubical cobar construction of $X$ to $G$. This confirms a result due to Minichiello-Rivera-Zeinalian and gives, as pointed out by these authors, a conceptual proof of the fact that the dga map $\Omega\,C(X) \to C(G)$ induced by Szczarba's twisting cochain is comultiplicative. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:51:35 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Franz",
"Matthias",
""
]
] |
2507.01670 | Joao Pires | Jo\~ao Pires | Precision determination of $\alpha_\text{s}$ from Dijet Cross Sections in the Multi-TeV Range | Contribution to the 2025 QCD session of the 59th Rencontres de Moriond | null | null | null | hep-ph hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this talk we present a determination of the strong coupling constant $\alpha_\text{s}$ and its energy-scale dependence based on a next-to-next-to-leading order (NNLO) QCD analysis of dijet production. Using the invariant mass of the dijet system to probe $\alpha_\text{s}$ at different scales, we extract a value of $\alpha_\text{s}(m_\text{Z})=0.1178 \pm 0.0022$ from LHC dijet data. The combination of various LHC datasets significantly extends the precision and scale reach of the analysis, enabling the first determination of $\alpha_\text{s}$ up to 7 TeV. By incorporating dijet cross sections from HERA, we further probe $\alpha_\text{s}$ at smaller scales, covering a kinematic range of more than three orders of magnitude. Our results are in excellent agreement with QCD predictions based on the renormalization group equation, providing a stringent test of the running of the strong coupling across a wide energy range. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:52:22 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Pires",
"João",
""
]
] |
2507.01671 | Jeremy Rekier | J. Rekier, S. A. Trian, A. Barik, D. Abdulah, W. Kang | Constraints on Earth's Core-Mantle boundary from nutation | Includes supplementary material as SI.pdf | null | null | null | astro-ph.EP physics.ao-ph physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | Periodic variations in the Sun and Moon's gravitational pull cause small changes in Earth's rotational axis direction called nutation. Nutation components in the retrograde quasi-diurnal frequency band measured in the terrestrial reference frame are amplified by resonance with the Free Core Nutation (FCN), a rotational mode of Earth's fluid core. Dissipative processes at the core-mantle boundary (CMB) dampen this resonance, contributing to the observed phase lag between tidal forcing and Earth's rotational response. This phase lag is commonly attributed to electromagnetic (EM) coupling between the core and the electrically conducting lower mantle. However, estimates of mantle conductivity and radial magnetic field strength at the CMB suggest these effects are insufficient. We show that the missing dissipation arises naturally from the excitation of internal waves in the fluid core by topographic features at the CMB. Adapting a theoretical framework originally developed for tidal flow over oceanic topography, we compute the form drag and associated power flux induced by CMB topography. Our results are consistent with a CMB topography characterized by a root mean square amplitude of a few kilometers. The model favors weak stratification at the top of the core, though stronger stratification remains compatible with increased topographic amplitude. This mechanism provides independent constraints on CMB topography and stratification, complementing seismological and magnetic observations. Its generality offers a new framework for probing deep-interior dynamics across terrestrial planets. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:52:33 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Rekier",
"J.",
""
],
[
"Trian",
"S. A.",
""
],
[
"Barik",
"A.",
""
],
[
"Abdulah",
"D.",
""
],
[
"Kang",
"W.",
""
]
] |
2507.01672 | Dmitrii Pavlov | Clemens Br\"user, Mario Kummer, Dmitrii Pavlov | Adjoints of Polytopes: Determinantal Representations and Smoothness | 31 pages, comments welcome | null | null | null | math.AG hep-th math.CO | http://creativecommons.org/licenses/by/4.0/ | In this article we study determinantal representations of adjoint hypersurfaces of polytopes. We prove that adjoint polynomials of all polygons can be represented as determinants of tridiagonal symmetric matrices of linear forms with the matrix size being equal to the degree of the adjoint. We prove a sufficient combinatorial condition for a surface in the projective three-space to have a determinantal representation and use it to show that adjoints of all three-dimensional polytopes with at most eight facets and a simple facet hyperplane arrangement admit a determinantal representation. This includes all such polytopes with a smooth adjoint. We demonstrate that, starting from four dimensions, adjoint hypersurfaces may not admit linear determinantal representations. Along the way we prove that, starting from three dimensions, adjoint hypersurfaces are typically singular, in contrast to the two-dimensional case. We also consider a special case of interest to physics, the ABHY associahedron. We construct a determinantal representation of its universal adjoint in three dimensions and show that in higher dimensions a similarly structured representation does not exist. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:54:41 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Brüser",
"Clemens",
""
],
[
"Kummer",
"Mario",
""
],
[
"Pavlov",
"Dmitrii",
""
]
] |
2507.01673 | Muzammil Behzad | Muzammil Behzad | Facial Emotion Learning with Text-Guided Multiview Fusion via Vision-Language Model for 3D/4D Facial Expression Recognition | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Facial expression recognition (FER) in 3D and 4D domains presents a significant challenge in affective computing due to the complexity of spatial and temporal facial dynamics. Its success is crucial for advancing applications in human behavior understanding, healthcare monitoring, and human-computer interaction. In this work, we propose FACET-VLM, a vision-language framework for 3D/4D FER that integrates multiview facial representation learning with semantic guidance from natural language prompts. FACET-VLM introduces three key components: Cross-View Semantic Aggregation (CVSA) for view-consistent fusion, Multiview Text-Guided Fusion (MTGF) for semantically aligned facial emotions, and a multiview consistency loss to enforce structural coherence across views. Our model achieves state-of-the-art accuracy across multiple benchmarks, including BU-3DFE, Bosphorus, BU-4DFE, and BP4D-Spontaneous. We further extend FACET-VLM to 4D micro-expression recognition (MER) on the 4DME dataset, demonstrating strong performance in capturing subtle, short-lived emotional cues. The extensive experimental results confirm the effectiveness and substantial contributions of each individual component within the framework. Overall, FACET-VLM offers a robust, extensible, and high-performing solution for multimodal FER in both posed and spontaneous settings. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:55:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Behzad",
"Muzammil",
""
]
] |
2507.01674 | Albachiara Cogo | Rodrigo Avalos, Albachiara Cogo, Andoni Royo Abrego | Conformal Green functions and Yamabe metrics of Sobolev regularity | 79 pages. Any comments are welcome | null | null | null | math.AP math.DG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We provide a full resolution of the Yamabe problem on closed 3-manifolds for Riemannian metrics of Sobolev class $W^{2,q}$ with $q > 3$. This requires developing an elliptic theory for the conformal Laplacian for rough metrics and establishing existence, regularity and a delicate blow-up analysis for its Green function. Most of the analytical work is carried out in dimensions $n \geq 3$ and for $W^{2,q}$ Riemannian metrics with $q>\tfrac{n}{2}$ and should be of independent interest. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:57:10 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Avalos",
"Rodrigo",
""
],
[
"Cogo",
"Albachiara",
""
],
[
"Abrego",
"Andoni Royo",
""
]
] |
2507.01675 | Zhicai Zhang | Yutao Zhu, Haoyang Fu, Wentai Luo, Shaomin Chen, Litao Yang, Zhicai Zhang | Extending the sensitivity of heavy sterile neutrino searches with solar neutrino experiments | 12 pages, 9 figures | null | null | null | hep-ph hep-ex | http://creativecommons.org/licenses/by/4.0/ | A sensitivity study of the search for heavy sterile neutrinos ($\nu_H$) in the MeV mass range using solar neutrino experiments is presented. $\nu_H$, with masses ranging from a few MeV up to 15 MeV, can be produced in the Sun through $^8$B decay and subsequently decay into $\nu_e e^+ e^-$. Its flux and lifetime strongly depend on the mixing parameter $|U_{eH}|^2$ and mass $m_{\nu_H}$. The $\nu_H$ signal can be detected via its decay products, either the $e^+e^-$ pair or $\nu_e$, depending on whether $\nu_H$ decays inside or outside the detector. Expected signal yields for both detection methods (detecting $e^+e^-$ or $\nu_e$ signal) are presented across the full $|U_{eH}|^2$ and $m_{\nu_H}$ parameter space. These two methods are found to be complementary in different regions of the $|U_{eH}|^2$ and $m_{\nu_H}$ phase space. By combining both approaches, we anticipate observing at least a handful of signal events across most of the parameter space of $10^{-6} < |U_{eH}|^2 < 1$ and 2 MeV $< m_{\nu_H} < $ 14 MeV, assuming a 500-ton solar neutrino experiment operating for one year. Key variables, such as the energy spectra of $\nu_e$ or $e^+e^-$ and the $\nu_e$ solar angle, are also discussed to help distinguish signal from major backgrounds, such as solar neutrino events. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 12:59:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhu",
"Yutao",
""
],
[
"Fu",
"Haoyang",
""
],
[
"Luo",
"Wentai",
""
],
[
"Chen",
"Shaomin",
""
],
[
"Yang",
"Litao",
""
],
[
"Zhang",
"Zhicai",
""
]
] |
2507.01676 | Giuseppe Ruggeri | Giuseppe Ruggeri, Renzo Andri, Daniele Jahier Pagliari, Lukas Cavigelli | Deep Recommender Models Inference: Automatic Asymmetric Data Flow Optimization | 5 pages, 4 figures, conference: IEEE ICCD24 | 2024 IEEE 42nd International Conference on Computer Design (ICCD), Milan, Italy, 2024, pp. 517-520 | 10.1109/ICCD63220.2024.00085 | null | cs.DC cs.AI cs.AR cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep Recommender Models (DLRMs) inference is a fundamental AI workload accounting for more than 79% of the total AI workload in Meta's data centers. DLRMs' performance bottleneck is found in the embedding layers, which perform many random memory accesses to retrieve small embedding vectors from tables of various sizes. We propose the design of tailored data flows to speedup embedding look-ups. Namely, we propose four strategies to look up an embedding table effectively on one core, and a framework to automatically map the tables asymmetrically to the multiple cores of a SoC. We assess the effectiveness of our method using the Huawei Ascend AI accelerators, comparing it with the default Ascend compiler, and we perform high-level comparisons with Nvidia A100. Results show a speed-up varying from 1.5x up to 6.5x for real workload distributions, and more than 20x for extremely unbalanced distributions. Furthermore, the method proves to be much more independent of the query distribution than the baseline. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:00:39 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ruggeri",
"Giuseppe",
""
],
[
"Andri",
"Renzo",
""
],
[
"Pagliari",
"Daniele Jahier",
""
],
[
"Cavigelli",
"Lukas",
""
]
] |
2507.01677 | John Pendry | JB Pendry, AJ Holden, WJ Stewart, I Youngs | Extremely Low Frequency Plasmons in Metallic Microstructures | null | Phys.Rev. Lett. 76 4773-6 (1996) | null | null | cond-mat.mes-hall | http://creativecommons.org/licenses/by/4.0/ | The plasmon is a well established collective excitation of metals in the visible and near UV but at much lower frequencies dissipation destroys all trace of the plasmon and typical Drude behaviour sets in. We propose a mechanism for depression of the plasma frequency into the far infra red or even GHz band: periodic structures build of very thin wires dilute the average concentration of electrons and considerably enhance the effective electron mass through self-inductance. Computations replicate the key features and confirm our analytic theory. The new structure has novel properties not observed before in the GHz band, including some possible impact on superconducting properties. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:00:50 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Pendry",
"JB",
""
],
[
"Holden",
"AJ",
""
],
[
"Stewart",
"WJ",
""
],
[
"Youngs",
"I",
""
]
] |
2507.01678 | Gabriele Barca | Gabriele Barca and Steffen Gielen | Bouncing Bianchi Models with Deformed Commutation Relations | 13 pages, 5 figures | null | null | null | gr-qc hep-th | http://creativecommons.org/licenses/by/4.0/ | We study the anisotropic Bianchi I and Bianchi II models in vacuum in the framework of deformed commutation relations (DCRs). Working in a parametrisation of the spatial metric by a volume and two anisotropy variables, we propose modified Poisson brackets that for the volume alone reproduce the bounce dynamics of effective loop quantum cosmology (LQC), with additional modifications for anisotropy degrees of freedom. We derive effective Friedmann equations and observe cosmological bounces both in Bianchi I and Bianchi II. For Bianchi II, we find that the cosmological bounce now interacts with the usual reflection seen in the Kasner indices in various interesting ways, in close similarity again with what had been seen in LQC. This suggests that the DCR framework could model more general quantum-gravity inspired bounce scenarios in a relatively straightforward way. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:00:57 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Barca",
"Gabriele",
""
],
[
"Gielen",
"Steffen",
""
]
] |
2507.01679 | Zeyu Huang | Zeyu Huang, Tianhao Cheng, Zihan Qiu, Zili Wang, Yinghui Xu, Edoardo M. Ponti, Ivan Titov | Blending Supervised and Reinforcement Fine-Tuning with Prefix Sampling | Work in progress | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing post-training techniques for large language models are broadly categorized into Supervised Fine-Tuning (SFT) and Reinforcement Fine-Tuning (RFT). Each paradigm presents a distinct trade-off: SFT excels at mimicking demonstration data but can lead to problematic generalization as a form of behavior cloning. Conversely, RFT can significantly enhance a model's performance but is prone to learn unexpected behaviors, and its performance is highly sensitive to the initial policy. In this paper, we propose a unified view of these methods and introduce Prefix-RFT, a hybrid approach that synergizes learning from both demonstration and exploration. Using mathematical reasoning problems as a testbed, we empirically demonstrate that Prefix-RFT is both simple and effective. It not only surpasses the performance of standalone SFT and RFT but also outperforms parallel mixed-policy RFT methods. A key advantage is its seamless integration into existing open-source frameworks, requiring only minimal modifications to the standard RFT pipeline. Our analysis highlights the complementary nature of SFT and RFT, and validates that Prefix-RFT effectively harmonizes these two learning paradigms. Furthermore, ablation studies confirm the method's robustness to variations in the quality and quantity of demonstration data. We hope this work offers a new perspective on LLM post-training, suggesting that a unified paradigm that judiciously integrates demonstration and exploration could be a promising direction for future research. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:04:09 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Huang",
"Zeyu",
""
],
[
"Cheng",
"Tianhao",
""
],
[
"Qiu",
"Zihan",
""
],
[
"Wang",
"Zili",
""
],
[
"Xu",
"Yinghui",
""
],
[
"Ponti",
"Edoardo M.",
""
],
[
"Titov",
"Ivan",
""
]
] |
2507.01680 | John Pendry | JB Pendry, AJ Holden, DJ Robbins, WJ Stewart | Magnetism from Conductors, and Enhanced Non-Linear Phenomena | null | IEEE 47, 2075-84 (1999) | null | null | physics.app-ph cond-mat.other | http://creativecommons.org/licenses/by/4.0/ | We show that microstructures built from non-magnetic conducting sheets exhibit an effective magnetic permeability, /mu_meff, which can be tuned to values not accessible in naturally occurring materials, including large imaginary components of /mu_meff. The microstructure is on a scale much less than the wavelength of radiation, is not resolved by incident microwaves, and uses a very low density of metal so that structures can be extremely lightweight. Most of the structures are resonant due to internal capacitance and inductance, and resonant enhancement combined with compression of electrical energy into a very small volume greatly enhances the energy density at critical locations in the structure, easily by factors of a million and possibly by much more. Weakly non-linear materials placed at these critical locations will show greatly enhanced effects raising the possibility of manufacturing active structures whose properties can be switched at will between many states | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:04:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Pendry",
"JB",
""
],
[
"Holden",
"AJ",
""
],
[
"Robbins",
"DJ",
""
],
[
"Stewart",
"WJ",
""
]
] |
2507.01681 | Nurgissa Yessirkegenov | Kuralay Apseit, Nurgissa Yessirkegenov, Amir Zhangirbayev | Sharp remainder of the $L^{p}$-Poincar\'e inequality for Baouendi-Grushin vector fields | 16 pages | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we establish a sharp remainder formula for the Poincar\'e inequality for Baouendi-Grushin vector fields in the setting of $L^{p}$ for complex-valued functions. In special cases, we recover previously known results. Additionally, we provide estimates of the remainder term for $p\geq2$ and $1<p<2\leq n<\infty$. As an application, we obtain a blow-up in finite time and global existence of the positive solutions to the initial-boundary value problem of the doubly nonlinear porous medium equation involving a degenerate nonlinear operator $\Delta_{\gamma,p}$. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:05:21 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Apseit",
"Kuralay",
""
],
[
"Yessirkegenov",
"Nurgissa",
""
],
[
"Zhangirbayev",
"Amir",
""
]
] |
2507.01682 | John Pendry | JB Pendry | Negative Refraction Makes a Perfect Lens | null | Phys. Rev. Lett. 85, 3966-9 (2000) | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | With a conventional lens sharpness of the image is always limited by the wavelength of light. An unconventional alternative to a lens, a slab of negative refractive index material, has the power to focus all Fourier components of a 2D image, even those that do not propagate in a radiative manner. Such super lenses can be realised in the microwave band with current technology and a version operating at the frequency of visible light, but at short distances of a few nanometres, can be realised in the form of a thin slab of silver as our simulations show. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:07:44 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Pendry",
"JB",
""
]
] |
2507.01683 | Marvin Bechtold | Marvin Bechtold, Johanna Barzen, Frank Leymann, Alexander Mandl | Simulating Quantum State Transfer between Distributed Devices using Noisy Interconnects | 38 pages, 6 figures | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scaling beyond individual quantum devices via distributed quantum computing relies critically on high-fidelity quantum state transfers between devices, yet the quantum interconnects needed for this are currently unavailable or expected to be significantly noisy. These limitations can be bypassed by simulating ideal state transfer using quasiprobability decompositions (QPDs). Wire cutting, for instance, allows this even without quantum interconnects. Nevertheless, QPD methods face drawbacks, requiring sampling from multiple circuit variants and incurring substantial sampling overhead. While prior theoretical work showed that incorporating noisy interconnects within QPD protocols could reduce sampling overhead relative to interconnect quality, a practical implementation for realistic conditions was lacking. Addressing this gap, this work presents a generalized and practical QPD for state transfer simulation using noisy interconnects to reduce sampling overhead. The QPD incorporates a single tunable parameter for straightforward calibration to any utilized interconnect. To lower practical costs, the work also explores reducing the number of distinct circuit variants required by the QPD. Experimental validation on contemporary quantum devices confirms the proposed QPD's practical feasibility and expected sampling overhead reduction under realistic noise. Notably, the results show higher effective state transfer fidelity than direct transfer over the underlying noisy interconnect. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:07:52 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bechtold",
"Marvin",
""
],
[
"Barzen",
"Johanna",
""
],
[
"Leymann",
"Frank",
""
],
[
"Mandl",
"Alexander",
""
]
] |
2507.01684 | Raymond Goldstein | Shantanu Raikwar, Adham Al-Kassem, Nir S. Gov, Adriana I. Pesci, Rapha\"el Jeanneret, Raymond E. Goldstein | Phototactic Decision-Making by Micro-Algae | 6 pages, 5 figures, with Supplementary Material appended | null | null | null | cond-mat.soft physics.bio-ph q-bio.CB | http://creativecommons.org/licenses/by/4.0/ | We study how simple eukaryotic organisms make decisions in response to competing stimuli in the context of phototaxis by the unicellular alga $Chlamydomonas~reinhardtii$. While negatively phototactic cells swim directly away from a collimated light beam, when presented with two beams of adjustable intersection angle and intensities, we find that cells swim in a direction given by an intensity-weighted average of the two light propagation vectors. This geometrical law is a fixed point of an adaptive model of phototaxis and minimizes the average light intensity falling on the anterior pole of the cell. At large angular separations, subpopulations of cells swim away from one source or the other, or along the direction of the geometrical law, with some cells stochastically switching between the three directions. This behavior is shown to arise from a population-level distribution of photoreceptor locations that breaks front-back symmetry of photoreception. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:10:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Raikwar",
"Shantanu",
""
],
[
"Al-Kassem",
"Adham",
""
],
[
"Gov",
"Nir S.",
""
],
[
"Pesci",
"Adriana I.",
""
],
[
"Jeanneret",
"Raphaël",
""
],
[
"Goldstein",
"Raymond E.",
""
]
] |
2507.01686 | Yu-Ming Yang | Yu-Ming Yang, Zhao-Chen Zhang, Xiao-Jun Bi, Peng-Fei Yin | Tidal Suppression of Fuzzy Dark Matter Heating in Milky Way Satellite Galaxies | 10 pages, 4 figures | null | null | null | astro-ph.CO astro-ph.GA hep-ph | http://creativecommons.org/licenses/by/4.0/ | Many previous studies have imposed stringent constraints on the particle mass of fuzzy dark matter (FDM) by analyzing observations of Galactic satellite galaxies, which show no significant evidence of the heating effect predicted by FDM. However, these analyses have generally neglected the tidal influence of the Milky Way, which can substantially suppress the FDM-induced heating in satellites. This oversight arises from computational challenges of accurately capturing the tidal effects in FDM simulations. In this study, we present a novel simulation framework that, for the first time, enables the simulation of an FDM-stellar system within an observationally motivated gravitational potential of the Milky Way. This framework incorporates the diverse Galactic components, including the gravitational influence of the Large Magellanic Cloud. Using the Fornax dwarf galaxy as a case study, we demonstrate that tidal effects significantly alleviate the tension between observational data and the predicted heating effect for an FDM particle mass of $m_a\sim 10^{-22}$ eV. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:11:56 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Yang",
"Yu-Ming",
""
],
[
"Zhang",
"Zhao-Chen",
""
],
[
"Bi",
"Xiao-Jun",
""
],
[
"Yin",
"Peng-Fei",
""
]
] |
2507.01687 | Stylianos Katsarakis | Georgios Arampatzis, Stylianos Katsarakis and Charalambos Makridakis | A generative modeling / Physics-Informed Neural Network approach to random differential equations | null | null | null | null | stat.ML cs.LG cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | The integration of Scientific Machine Learning (SciML) techniques with uncertainty quantification (UQ) represents a rapidly evolving frontier in computational science. This work advances Physics-Informed Neural Networks (PINNs) by incorporating probabilistic frameworks to effectively model uncertainty in complex systems. Our approach enhances the representation of uncertainty in forward problems by combining generative modeling techniques with PINNs. This integration enables in a systematic fashion uncertainty control while maintaining the predictive accuracy of the model. We demonstrate the utility of this method through applications to random differential equations and random partial differential equations (PDEs). | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:14:17 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Arampatzis",
"Georgios",
""
],
[
"Katsarakis",
"Stylianos",
""
],
[
"Makridakis",
"Charalambos",
""
]
] |
2507.01688 | Mikel Palmero | Mikel Palmero, Juan Gonzalo Muga, Ander Tobalina | Energetically efficient, mediated mechanical system for precise control of hoisting operations | 5 pages, 3 figures, 1 table | null | null | null | physics.class-ph quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a mechanical control system for energy efficient and robust hoisting crane operations. The control system efficiently translates the harmonic motion of a spring loaded mediating system into the desired driving of the load, recyling most of the employed energy for subsequent operations. The control output is a shortcut-to-adiabaticity protocol borrowed from quantum mechanics. The control system reduces the single operation consuption in realistic working regimes, but it is in cyclical processes where the energetical advantage becomes substantial. The design of the control system and the control output is flexible enough to allow additional optimization of the robustness against perturbations. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:14:45 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Palmero",
"Mikel",
""
],
[
"Muga",
"Juan Gonzalo",
""
],
[
"Tobalina",
"Ander",
""
]
] |
2507.01689 | Nikolaos Mavromatos | Panagiotis Dorlis, Nick E. Mavromatos, Sarben Sarkar, Sotirios-Neilos Vlachos | Superradiant Axionic Black-Hole Clouds as Seeds for Graviton Squeezing | 7 pages revtex, four pdf figures incorporated | null | null | KCL-PH-TH/2025-15 | gr-qc hep-ph hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is shown that both, standard general relativity (GR) and Chern-Simons (CS) gravity, the latter containing chiral gravitational anomaly terms, seed the production of pairs of entangled gravitons in a multi-mode squeezed state. This involves the interaction of gravitons with the axionic cloud surrounding a superradiant Kerr (rotating) black hole background. The order of magnitude of the squeezing effect, specifically the number of graviton excitations in the squeezed vacuum, is estimated in the non-relativistic limit, relevant for the superradiance process. It is found analytically that the squeezing from the GR process of annihilation of two axions into two gravitons, dominates, by many orders of magnitude, that coming from the axion decay into two gravitons, induced by the higher-derivative CS term. It is also shown that significant squeezing effects are produced in the case of long-lived axionic clouds, whose lifetimes are much longer than the timescale for which superradiance is effective. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:17:22 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Dorlis",
"Panagiotis",
""
],
[
"Mavromatos",
"Nick E.",
""
],
[
"Sarkar",
"Sarben",
""
],
[
"Vlachos",
"Sotirios-Neilos",
""
]
] |
2507.01690 | Ana Henriques | Beatriz Severes and Ana O. Henriques and Rory Clark and Paulo Bala and Anna Carter and Rua Mae Williams and Geraldine Fitzpatrick | Designing for Community Care: Reimagining Support for Equity & Well-being in Academia | null | null | 10.1145/3715668.3734177 | null | cs.HC | http://creativecommons.org/licenses/by-sa/4.0/ | Academic well-being is deeply influenced by peer-support networks, yet they remain informal, inequitable, and unsustainable, often relying on personal connections and social capital rather than structured, inclusive systems. Additionally, institutional well-being responses frequently focus on student populations, neglecting the emotional labour of faculty and staff, reinforcing an exclusionary academic culture. Drawing on HCI methodologies, participatory design, and care ethics, this workshop will provide a space for rethinking how academic communities can support inclusive networks. Through pre-workshop engagement, co-design activities, and reflection, participants will examine systemic gaps in networks and explore ways to embed care, equity, and sustainability into academic peer-support frameworks -- from informal, exclusionary models to structured, inclusive care-based ecosystems. At the end of the workshop, participants will co-develop design strategies for integrating care and resilience in academic ecosystems, resources for designing equitable support systems, and a peer network invested and committed to fostering a supportive academic community. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:17:38 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Severes",
"Beatriz",
""
],
[
"Henriques",
"Ana O.",
""
],
[
"Clark",
"Rory",
""
],
[
"Bala",
"Paulo",
""
],
[
"Carter",
"Anna",
""
],
[
"Williams",
"Rua Mae",
""
],
[
"Fitzpatrick",
"Geraldine",
""
]
] |
2507.01691 | Oliver Sefrin | Oliver Sefrin, Manuel Radons, Lars Simon, Sabine W\"olk | Quantum reinforcement learning in dynamic environments | 22 pages, 8 figures, submitted to Quantum Machine Intelligence | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Combining quantum computing techniques in the form of amplitude amplification with classical reinforcement learning has led to the so-called "hybrid agent for quantum-accessible reinforcement learning", which achieves a quadratic speedup in sample complexity for certain learning problems. So far, this hybrid agent has only been applied to stationary learning problems, that is, learning problems without any time dependency within components of the Markov decision process. In this work, we investigate the applicability of the hybrid agent to dynamic RL environments. To this end, we enhance the hybrid agent by introducing a dissipation mechanism and, with the resulting learning agent, perform an empirical comparison with a classical RL agent in an RL environment with a time-dependent reward function. Our findings suggest that the modified hybrid agent can adapt its behavior to changes in the environment quickly, leading to a higher average success probability compared to its classical counterpart. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:17:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Sefrin",
"Oliver",
""
],
[
"Radons",
"Manuel",
""
],
[
"Simon",
"Lars",
""
],
[
"Wölk",
"Sabine",
""
]
] |
2507.01692 | Silius M. Vandeskog | Silius M. Vandeskog and Thordis L. Thorarinsdottir and Alex Lenkoski | Simulation and evaluation of local daily temperature and precipitation series derived by stochastic downscaling of ERA5 reanalysis | null | null | null | null | stat.AP | http://creativecommons.org/licenses/by/4.0/ | Reanalysis products such as the ERA5 reanalysis are commonly used as proxies for observed atmospheric conditions. These products are convenient to use due to their global coverage, the large number of available atmospheric variables and the physical consistency between these variables, as well as their relatively high spatial and temporal resolutions. However, despite the continuous improvements in accuracy and increasing spatial and temporal resolutions of reanalysis products, they may not always capture local atmospheric conditions, especially for highly localised variables such as precipitation. This paper proposes a computationally efficient stochastic downscaling of ERA5 temperature and precipitation. The method combines information from ERA5 and surface observations from nearby stations in a non-linear regression framework that combines generalised additive models (GAMs) with regression splines and auto-regressive moving average (ARMA) models to produce realistic time series of local daily temperature and precipitation. Using a wide range of evaluation criteria that address different properties of the data, the proposed framework is shown to improve the representation of local temperature and precipitation compared to ERA5 at over 4000 locations in Europe over a period of 60 years. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:19:51 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Vandeskog",
"Silius M.",
""
],
[
"Thorarinsdottir",
"Thordis L.",
""
],
[
"Lenkoski",
"Alex",
""
]
] |
2507.01693 | Adrians Skapars | Adrians Skapars, Edoardo Manino, Youcheng Sun, Lucas C. Cordeiro | GPT, But Backwards: Exactly Inverting Language Model Outputs | 9 pages, ICML 2025 Workshop on Reliable and Responsible Foundation Models | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While existing auditing techniques attempt to identify potential unwanted behaviours in large language models (LLMs), we address the complementary forensic problem of reconstructing the exact input that led to an existing LLM output - enabling post-incident analysis and potentially the detection of fake output reports. We formalize exact input reconstruction as a discrete optimisation problem with a unique global minimum and introduce SODA, an efficient gradient-based algorithm that operates on a continuous relaxation of the input search space with periodic restarts and parameter decay. Through comprehensive experiments on LLMs ranging in size from 33M to 3B parameters, we demonstrate that SODA significantly outperforms existing approaches. We succeed in fully recovering 79.5% of shorter out-of-distribution inputs from next-token logits, without a single false positive, but struggle to extract private information from the outputs of longer (15+ token) input sequences. This suggests that standard deployment practices may currently provide adequate protection against malicious use of our method. Our code is available at https://doi.org/10.5281/zenodo.15539879. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:20:30 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Skapars",
"Adrians",
""
],
[
"Manino",
"Edoardo",
""
],
[
"Sun",
"Youcheng",
""
],
[
"Cordeiro",
"Lucas C.",
""
]
] |
2507.01694 | Hanlin Cai | Hanlin Cai, Haofan Dong, Houtianfu Wang, Kai Li, Ozgur B. Akan | Graph Representation-based Model Poisoning on Federated LLMs in CyberEdge Networks | 7 pages, 5 figures | null | null | null | cs.CR cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated large language models (FedLLMs) provide powerful generative capabilities in CyberEdge networks while protecting data privacy. However, FedLLMs remains highly vulnerable to model poisoning attacks. This article first reviews recent model poisoning techniques and existing defense mechanisms for FedLLMs, highlighting critical limitations, particularly under non-IID text distributions. In particular, current defenses primarily utilize distance-based outlier detection or norm constraints, operating under the assumption that adversarial updates significantly diverge from benign statistics. This assumption can fail when facing adaptive attackers targeting billionparameter LLMs. Next, this article investigates emerging Graph Representation-Based Model Poisoning (GRMP), a novel attack paradigm that leverages higher-order correlations among honest client gradients to synthesize malicious updates indistinguishable from legitimate model updates. GRMP can effectively evade advanced defenses, resulting in substantial accuracy loss and performance degradation. Moreover, this article outlines a research roadmap emphasizing the importance of graph-aware secure aggregation methods, FedLLMs-specific vulnerability metrics, and evaluation frameworks to strengthen the robustness of future federated language model deployments. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:20:52 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Cai",
"Hanlin",
""
],
[
"Dong",
"Haofan",
""
],
[
"Wang",
"Houtianfu",
""
],
[
"Li",
"Kai",
""
],
[
"Akan",
"Ozgur B.",
""
]
] |
2507.01695 | Marcello Traiola | Omkar Shende, Gayathri Ananthanarayanan, Marcello Traiola | PERTINENCE: Input-based Opportunistic Neural Network Dynamic Execution | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Deep neural networks (DNNs) have become ubiquitous thanks to their remarkable ability to model complex patterns across various domains such as computer vision, speech recognition, robotics, etc. While large DNN models are often more accurate than simpler, lightweight models, they are also resource- and energy-hungry. Hence, it is imperative to design methods to reduce reliance on such large models without significant degradation in output accuracy. The high computational cost of these models is often necessary only for a reduced set of challenging inputs, while lighter models can handle most simple ones. Thus, carefully combining properties of existing DNN models in a dynamic, input-based way opens opportunities to improve efficiency without impacting accuracy.
In this work, we introduce PERTINENCE, a novel online method designed to analyze the complexity of input features and dynamically select the most suitable model from a pre-trained set to process a given input effectively. To achieve this, we employ a genetic algorithm to explore the training space of an ML-based input dispatcher, enabling convergence towards the Pareto front in the solution space that balances overall accuracy and computational efficiency.
We showcase our approach on state-of-the-art Convolutional Neural Networks (CNNs) trained on the CIFAR-10 and CIFAR-100, as well as Vision Transformers (ViTs) trained on TinyImageNet dataset. We report results showing PERTINENCE's ability to provide alternative solutions to existing state-of-the-art models in terms of trade-offs between accuracy and number of operations. By opportunistically selecting among models trained for the same task, PERTINENCE achieves better or comparable accuracy with up to 36% fewer operations. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:22:05 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Shende",
"Omkar",
""
],
[
"Ananthanarayanan",
"Gayathri",
""
],
[
"Traiola",
"Marcello",
""
]
] |
2507.01696 | Peter Macgregor | Steinar Laenen, Peter Macgregor, He Sun | Dynamic Similarity Graph Construction with Kernel Density Estimation | ICML'25 | null | null | null | cs.DS cs.LG | http://creativecommons.org/licenses/by/4.0/ | In the kernel density estimation (KDE) problem, we are given a set $X$ of data points in $\mathbb{R}^d$, a kernel function $k: \mathbb{R}^d \times \mathbb{R}^d \rightarrow \mathbb{R}$, and a query point $\mathbf{q} \in \mathbb{R}^d$, and the objective is to quickly output an estimate of $\sum_{\mathbf{x} \in X} k(\mathbf{q}, \mathbf{x})$. In this paper, we consider $\textsf{KDE}$ in the dynamic setting, and introduce a data structure that efficiently maintains the estimates for a set of query points as data points are added to $X$ over time. Based on this, we design a dynamic data structure that maintains a sparse approximation of the fully connected similarity graph on $X$, and develop a fast dynamic spectral clustering algorithm. We further evaluate the effectiveness of our algorithms on both synthetic and real-world datasets. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:25:22 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Laenen",
"Steinar",
""
],
[
"Macgregor",
"Peter",
""
],
[
"Sun",
"He",
""
]
] |
2507.01697 | Xiao-Song Yang | Yu Zhang, Qi Zhou, Xiao-Song Yang | An RRT* algorithm based on Riemannian metric model for optimal path planning | 27 pages | null | null | null | cs.RO math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a Riemannian metric-based model to solve the optimal path planning problem on two-dimensional smooth submanifolds in high-dimensional space. Our model is based on constructing a new Riemannian metric on a two-dimensional projection plane, which is induced by the high-dimensional Euclidean metric on two-dimensional smooth submanifold and reflects the environmental information of the robot. The optimal path planning problem in high-dimensional space is therefore transformed into a geometric problem on the two-dimensional plane with new Riemannian metric. Based on the new Riemannian metric, we proposed an incremental algorithm RRT*-R on the projection plane. The experimental results show that the proposed algorithm is suitable for scenarios with uneven fields in multiple dimensions. The proposed algorithm can help the robot to effectively avoid areas with drastic changes in height, ground resistance and other environmental factors. More importantly, the RRT*-R algorithm shows better smoothness and optimization properties compared with the original RRT* algorithm using Euclidean distance in high-dimensional workspace. The length of the entire path by RRT*-R is a good approximation of the theoretical minimum geodesic distance on projection plane. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:26:33 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Zhang",
"Yu",
""
],
[
"Zhou",
"Qi",
""
],
[
"Yang",
"Xiao-Song",
""
]
] |
2507.01698 | Rocco D'Agostino | Rocco D'Agostino, Vittorio De Falco | Quasinormal modes of nonlocal gravity black holes | 13 pages, 5 figures | null | null | null | gr-qc astro-ph.HE hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a comprehensive study of the quasinormal modes of a new class of nonlocal static and spherically symmetric black hole (BH) solutions within the framework of the revised Deser-Woodard theory of gravity. These solutions are constructed as linear perturbations of the Schwarzschild spacetime and are characterized by an inverse power-law behavior of the lapse metric function. We derive the radial profiles of the effective potentials corresponding to scalar, electromagnetic and axial gravitational fluctuations on the BH background. Using the WKB method, complemented by Pad\'e approximants to regularize the trend of the effective potential near its peak, we compute the complex quasinormal mode frequencies associated with each type of perturbation. Our results are examined from both mathematical and physical perspectives, and are substantiated with references to existing literature. In particular, we compare the numerical outcomes with the predictions of the Schwarzschild metric to quantify deviations from the framework of general relativity. When all types of perturbations are combined, the relative deviations of the fundamental modes can reach up to $\sim 12\%$. Finally, we discuss the possibility to place observational bounds in the BH parameter space, based on the predicted sensitivities of future gravitational wave detectors. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:28:11 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"D'Agostino",
"Rocco",
""
],
[
"De Falco",
"Vittorio",
""
]
] |
2507.01699 | Illia Oleksiienko | Illia Oleksiienko, Juho Kanniainen, and Alexandros Iosifidis | Variational Graph Convolutional Neural Networks | This work has been submitted to the IEEE for possible publication. 9 pages, 6 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Estimation of model uncertainty can help improve the explainability of Graph Convolutional Networks and the accuracy of the models at the same time. Uncertainty can also be used in critical applications to verify the results of the model by an expert or additional models. In this paper, we propose Variational Neural Network versions of spatial and spatio-temporal Graph Convolutional Networks. We estimate uncertainty in both outputs and layer-wise attentions of the models, which has the potential for improving model explainability. We showcase the benefits of these models in the social trading analysis and the skeleton-based human action recognition tasks on the Finnish board membership, NTU-60, NTU-120 and Kinetics datasets, where we show improvement in model accuracy in addition to estimated model uncertainties. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:28:37 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Oleksiienko",
"Illia",
""
],
[
"Kanniainen",
"Juho",
""
],
[
"Iosifidis",
"Alexandros",
""
]
] |
2507.01700 | Andrea Piras | Andrea Piras, Matteo Negro, Ragib Ahsan, David Arbour, Elena Zheleva | Relational Causal Discovery with Latent Confounders | 30 pages, 19 figures. Accepted for publication at the 41st Conference on Uncertainty in Artificial Intelligence (UAI 2025). Andrea Piras and Matteo Negro contributed equally to this work | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Estimating causal effects from real-world relational data can be challenging when the underlying causal model and potential confounders are unknown. While several causal discovery algorithms exist for learning causal models with latent confounders from data, they assume that the data is independent and identically distributed (i.i.d.) and are not well-suited for learning from relational data. Similarly, existing relational causal discovery algorithms assume causal sufficiency, which is unrealistic for many real-world datasets. To address this gap, we propose RelFCI, a sound and complete causal discovery algorithm for relational data with latent confounders. Our work builds upon the Fast Causal Inference (FCI) and Relational Causal Discovery (RCD) algorithms and it defines new graphical models, necessary to support causal discovery in relational domains. We also establish soundness and completeness guarantees for relational d-separation with latent confounders. We present experimental results demonstrating the effectiveness of RelFCI in identifying the correct causal structure in relational causal models with latent confounders. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:29:35 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Piras",
"Andrea",
""
],
[
"Negro",
"Matteo",
""
],
[
"Ahsan",
"Ragib",
""
],
[
"Arbour",
"David",
""
],
[
"Zheleva",
"Elena",
""
]
] |
2507.01701 | Bochen Han | Bochen Han, Songmao Zhang | Exploring Advanced LLM Multi-Agent Systems Based on Blackboard Architecture | null | null | null | null | cs.MA cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose to incorporate the blackboard architecture into LLM multi-agent systems (MASs) so that (1) agents with various roles can share all the information and others' messages during the whole problem-solving process, (2) agents that will take actions are selected based on the current content of the blackboard, and (3) the selection and execution round is repeated until a consensus is reached on the blackboard. We develop the first implementation of this proposal and conduct experiments on commonsense knowledge, reasoning and mathematical datasets. The results show that our system can be competitive with the SOTA static and dynamic MASs by achieving the best average performance, and at the same time manage to spend less tokens. Our proposal has the potential to enable complex and dynamic problem-solving where well-defined structures or workflows are unavailable. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:30:44 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Han",
"Bochen",
""
],
[
"Zhang",
"Songmao",
""
]
] |
2507.01702 | Lin Hongzhan | Zixin Chen, Hongzhan Lin, Kaixin Li, Ziyang Luo, Zhen Ye, Guang Chen, Zhiyong Huang, Jing Ma | AdamMeme: Adaptively Probe the Reasoning Capacity of Multimodal Large Language Models on Harmfulness | ACL 2025 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The proliferation of multimodal memes in the social media era demands that multimodal Large Language Models (mLLMs) effectively understand meme harmfulness. Existing benchmarks for assessing mLLMs on harmful meme understanding rely on accuracy-based, model-agnostic evaluations using static datasets. These benchmarks are limited in their ability to provide up-to-date and thorough assessments, as online memes evolve dynamically. To address this, we propose AdamMeme, a flexible, agent-based evaluation framework that adaptively probes the reasoning capabilities of mLLMs in deciphering meme harmfulness. Through multi-agent collaboration, AdamMeme provides comprehensive evaluations by iteratively updating the meme data with challenging samples, thereby exposing specific limitations in how mLLMs interpret harmfulness. Extensive experiments show that our framework systematically reveals the varying performance of different target mLLMs, offering in-depth, fine-grained analyses of model-specific weaknesses. Our code is available at https://github.com/Lbotirx/AdamMeme. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:32:30 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Chen",
"Zixin",
""
],
[
"Lin",
"Hongzhan",
""
],
[
"Li",
"Kaixin",
""
],
[
"Luo",
"Ziyang",
""
],
[
"Ye",
"Zhen",
""
],
[
"Chen",
"Guang",
""
],
[
"Huang",
"Zhiyong",
""
],
[
"Ma",
"Jing",
""
]
] |
2507.01703 | Prasanta Gorai | Prasanta Gorai, Kotomi Taniguchi, Jonathan C. Tan, Miguel Gomez-Garrido, Viviana Rosero, Izaskun Jimenez-Serra, Yichen Zhang, Giuliana Cosentino, Chi-Yan Law, Ruben Fedriani, Gemma Busquet, Brandt A. L. Gaches, Maryam Saberi, Ankan Das | The SOFIA Massive (SOMA) Star Formation Q-band Follow-up. II. Hydrogen Recombination Lines Toward High-Mass Protostars | submitted to A&A, comments welcome | null | null | null | astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Hydrogen recombination lines (HRLs) are valuable diagnostics of the physical conditions in ionized regions around high-mass stars. Understanding their broadening mechanisms and intensity trends can provide insights into the densities, temperatures, and kinematics of HII regions. We investigate the properties of ionized gas around massive protostars by analyzing hydrogen recombination lines (H-alpha and H-beta) in the Q-band. Observations were conducted using the Yebes 40m radio telescope in the Q-band (30.5~50 GHz) toward six high-mass protostars selected from the SOMA Survey (G45.12+0.13, G45.47+0.05, G28.20-0.05, G35.20-0.74, G19.08-0.29, and G31.28+0.06). The line profiles were analyzed to assess broadening mechanisms, from which electron densities and temperatures were derived. We compared our results with Q-band data from the TianMa 65m Radio Telescope (TMRT) and ALMA Band 1 Science Verification observations of Orion KL. A total of eight H-alpha (n = 51 to 58) and ten H-beta (n = 64 to 73) lines were detected toward G45.12+0.13, G45.47+0.05, and G28.20-0.05, with non-detections in the other sources. Electron densities of \sim \mathrm{1\text{--}5 \times 10^{6}~cm^{-3}} and temperatures of 8000-10000 K were derived. Orion KL shows one order of magnitude lower electron density, but a similar temperature. Notably, G45.12 and G28.20 show increasing intensity with frequency for both H-alpha and H-beta, in contrast to the decreasing trend in Orion KL. The observed line widths indicate contributions from both thermal and dynamical broadening, suggesting high-temperature ionized gas affected by turbulence, outflows, rotation, or stellar winds. Pressure broadening may also play a minor role. The contrasting intensity trends likely reflect differences in local physical conditions or radiative transfer effects, warranting further study through higher-resolution observations and modeling. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:36:27 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gorai",
"Prasanta",
""
],
[
"Taniguchi",
"Kotomi",
""
],
[
"Tan",
"Jonathan C.",
""
],
[
"Gomez-Garrido",
"Miguel",
""
],
[
"Rosero",
"Viviana",
""
],
[
"Jimenez-Serra",
"Izaskun",
""
],
[
"Zhang",
"Yichen",
""
],
[
"Cosentino",
"Giuliana",
""
],
[
"Law",
"Chi-Yan",
""
],
[
"Fedriani",
"Ruben",
""
],
[
"Busquet",
"Gemma",
""
],
[
"Gaches",
"Brandt A. L.",
""
],
[
"Saberi",
"Maryam",
""
],
[
"Das",
"Ankan",
""
]
] |
2507.01704 | Oliver Surbek | Felix K\"ubler, Simon Scheidegger, Oliver Surbek | Using Machine Learning to Compute Constrained Optimal Carbon Tax Rules | null | null | null | null | econ.GN q-fin.EC | http://creativecommons.org/licenses/by/4.0/ | We develop a computational framework for deriving Pareto-improving and constrained optimal carbon tax rules in a stochastic overlapping generations (OLG) model with climate change. By integrating Deep Equilibrium Networks for fast policy evaluation and Gaussian process surrogate modeling with Bayesian active learning, the framework systematically locates optimal carbon tax schedules for heterogeneous agents exposed to climate risk. We apply our method to a 12-period OLG model in which exogenous shocks affect the carbon intensity of energy production, as well as the damage function. Constrained optimal carbon taxes consist of tax rates that are simple functions of observables and revenue-sharing rules that guarantee that the introduction of the taxes is Pareto improving. This reveals that a straightforward policy is highly effective: a Pareto-improving linear tax on cumulative emissions alone yields a 0.42% aggregate welfare gain in consumption-equivalent terms while adding further complexity to the tax provides only a marginal increase to 0.45%. The application demonstrates that the proposed approach produces scalable tools for macro-policy design in complex stochastic settings. Beyond climate economics, the framework offers a template for systematically analyzing welfare-improving policies in various heterogeneous-agent problems. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:36:46 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Kübler",
"Felix",
""
],
[
"Scheidegger",
"Simon",
""
],
[
"Surbek",
"Oliver",
""
]
] |
2507.01705 | Marc-Philip Ecker | Marc-Philip Ecker, Bernhard Bischof, Minh Nhat Vu, Christoph Fr\"ohlich, Tobias Gl\"uck, Wolfgang Kemmetm\"uller | Efficient Collision Detection for Long and Slender Robotic Links in Euclidean Distance Fields: Application to a Forestry Crane | Accepted at IROS 2025 | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Collision-free motion planning in complex outdoor environments relies heavily on perceiving the surroundings through exteroceptive sensors. A widely used approach represents the environment as a voxelized Euclidean distance field, where robots are typically approximated by spheres. However, for large-scale manipulators such as forestry cranes, which feature long and slender links, this conventional spherical approximation becomes inefficient and inaccurate. This work presents a novel collision detection algorithm specifically designed to exploit the elongated structure of such manipulators, significantly enhancing the computational efficiency of motion planning algorithms. Unlike traditional sphere decomposition methods, our approach not only improves computational efficiency but also naturally eliminates the need to fine-tune the approximation accuracy as an additional parameter. We validate the algorithm's effectiveness using real-world LiDAR data from a forestry crane application, as well as simulated environment data. | [
{
"version": "v1",
"created": "Wed, 02 Jul 2025 13:37:08 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Ecker",
"Marc-Philip",
""
],
[
"Bischof",
"Bernhard",
""
],
[
"Vu",
"Minh Nhat",
""
],
[
"Fröhlich",
"Christoph",
""
],
[
"Glück",
"Tobias",
""
],
[
"Kemmetmüller",
"Wolfgang",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.