id
string | submitter
string | authors
string | title
string | comments
string | journal-ref
string | doi
string | report-no
string | categories
string | license
string | abstract
string | versions
list | update_date
timestamp[s] | authors_parsed
sequence |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2507.02780 | Claude Cournoyer-Cloutier | Claude Cournoyer-Cloutier, Eric P. Andersson, Sabrina M. Appel, Natalia Lah\'en, Brooke Polak, Antti Rantala, Silvia Toonen, Alison Sills, Steven Rieder, Simon Portegies Zwart, Mordecai-Mark Mac Low, William E. Harris | Massive Interacting Binaries Enhance Feedback in Star-Forming Regions | Accepted for publication in the Astrophysical Journal. 28 pages (including references), 12 figures, 2 tables | null | null | null | astro-ph.GA astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a new framework to incorporate feedback from massive interacting binaries in simulations of star cluster formation. Our new feedback model adds binary stellar evolution to the cluster formation code Torch, and couples it in AMUSE to the pre-existing modules for collisional stellar dynamics, magnetohydrodynamics, and mechanical and radiative feedback. Our model accounts for the effects of mass transfer on the stars' mass loss rates, their radiation spectra, and the timing of core-collapse supernovae. It also injects mass lost through non-conservative mass transfer and common envelope ejection into the interstellar medium. We demonstrate the use of our feedback model through simulations of isolated binaries in a gaseous medium, and of embedded clusters of massive binaries. Feedback from interacting binaries efficiently couples with the surrounding interstellar medium. It increases the size of HII regions, increases the kinetic and thermal energy of the gas, and increases the pressure within HII regions compared to models that use single star stellar evolution. Those differences arise from the ionizing radiation, which increases by three orders of magnitude, resulting in HII regions that expand due to thermal pressure rather than radiation pressure. The effects of stellar dynamics and the gravitational potential of the background gas cause the evolution of individual binaries to deviate from the predictions made by secular evolution, impacting the subsequent feedback from the binary. We conclude that massive interacting binaries are an important source of feedback in cluster-forming regions, and must be considered when studying the emerging timescales of young star clusters. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:43:11 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Cournoyer-Cloutier",
"Claude",
""
],
[
"Andersson",
"Eric P.",
""
],
[
"Appel",
"Sabrina M.",
""
],
[
"Lahén",
"Natalia",
""
],
[
"Polak",
"Brooke",
""
],
[
"Rantala",
"Antti",
""
],
[
"Toonen",
"Silvia",
""
],
[
"Sills",
"Alison",
""
],
[
"Rieder",
"Steven",
""
],
[
"Zwart",
"Simon Portegies",
""
],
[
"Mac Low",
"Mordecai-Mark",
""
],
[
"Harris",
"William E.",
""
]
] |
2507.02781 | Huili Huang | Danrong Zhang, Huili Huang, N. Simrill Smith, Nimisha Roy, J. David Frost | From Pixels to Damage Severity: Estimating Earthquake Impacts Using Semantic Segmentation of Social Media Images | null | null | null | null | cs.CV cs.SI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In the aftermath of earthquakes, social media images have become a crucial resource for disaster reconnaissance, providing immediate insights into the extent of damage. Traditional approaches to damage severity assessment in post-earthquake social media images often rely on classification methods, which are inherently subjective and incapable of accounting for the varying extents of damage within an image. Addressing these limitations, this study proposes a novel approach by framing damage severity assessment as a semantic segmentation problem, aiming for a more objective analysis of damage in earthquake-affected areas. The methodology involves the construction of a segmented damage severity dataset, categorizing damage into three degrees: undamaged structures, damaged structures, and debris. Utilizing this dataset, the study fine-tunes a SegFormer model to generate damage severity segmentations for post-earthquake social media images. Furthermore, a new damage severity scoring system is introduced, quantifying damage by considering the varying degrees of damage across different areas within images, adjusted for depth estimation. The application of this approach allows for the quantification of damage severity in social media images in a more objective and comprehensive manner. By providing a nuanced understanding of damage, this study enhances the ability to offer precise guidance to disaster reconnaissance teams, facilitating more effective and targeted response efforts in the aftermath of earthquakes. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:43:56 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Danrong",
""
],
[
"Huang",
"Huili",
""
],
[
"Smith",
"N. Simrill",
""
],
[
"Roy",
"Nimisha",
""
],
[
"Frost",
"J. David",
""
]
] |
2507.02782 | Ricardo Buitrago Ruiz | Ricardo Buitrago Ruiz and Albert Gu | Understanding and Improving Length Generalization in Recurrent Models | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recently, recurrent models such as state space models and linear attention have become popular due to their linear complexity in the sequence length. Thanks to their recurrent nature, in principle they can process arbitrarily long sequences, but their performance sometimes drops considerably beyond their training context lengths-i.e. they fail to length generalize. In this work, we provide comprehensive empirical and theoretical analysis to support the unexplored states hypothesis, which posits that models fail to length generalize when during training they are only exposed to a limited subset of the distribution of all attainable states (i.e. states that would be attained if the recurrence was applied to long sequences). Furthermore, we investigate simple training interventions that aim to increase the coverage of the states that the model is trained on, e.g. by initializing the state with Gaussian noise or with the final state of a different input sequence. With only 500 post-training steps ($\sim 0.1\%$ of the pre-training budget), these interventions enable length generalization for sequences that are orders of magnitude longer than the training context (e.g. $2k\longrightarrow 128k$) and show improved performance in long context tasks, thus presenting a simple and efficient way to enable robust length generalization in general recurrent models. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:45:50 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ruiz",
"Ricardo Buitrago",
""
],
[
"Gu",
"Albert",
""
]
] |
2507.02783 | Conrad Qu | Di Fang, Conrad Qu | Uniform semiclassical observable error bound of Trotterization without the Egorov theorem: a simple algebraic proof | null | null | null | null | math.NA cs.NA quant-ph | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Efficient simulation of the semiclassical Schr\"odinger equation has garnered significant attention in the numerical analysis community. While controlling the error in the unitary evolution or the wavefunction typically requires the time step size to shrink as the semiclassical parameter $h$ decreases, it has been observed -- and proved for first- and second-order Trotterization schemes -- that the error in certain classes of observables admits a time step size independent of $h$. In this work, we explicitly characterize this class of observables and present a new, simple algebraic proof of uniform-in-$h$ error bounds for arbitrarily high-order Trotterization schemes. Our proof relies solely on the algebraic structure of the underlying operators in both the continuous and discrete settings. Unlike previous analyses, it avoids Egorov-type theorems and bypasses heavy semiclassical machinery. To our knowledge, this is the first proof of uniform-in-$h$ observable error bounds for Trotterization in the semiclassical regime that relies only on algebraic structure, without invoking the semiclassical limit. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:47:29 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fang",
"Di",
""
],
[
"Qu",
"Conrad",
""
]
] |
2507.02784 | Benjamin Vial | Benjamin Vial and Richard V. Craster | Quasinormal modes of Floquet media slabs | null | null | null | null | physics.optics | http://creativecommons.org/licenses/by/4.0/ | Exploiting non-Hermitian wave-matter interactions in time-modulated media to enable the dynamic control of electromagnetic waves requires advanced theoretical tools. In this article we bridge concepts from photonic quasinormal modes (QNMs) and time-varying metamaterials providing the foundation for designing dynamic optical devices with prescribed scattering properties. Establishing the QNM framework for slabs with time-periodic permittivity, and solving the associated nonlinear eigenvalue problem, allows us to derive the QNM expansion capturing the resonant features of the system. This reduced-order model enables highly efficient computation of scattered fields while revealing insight into how modulation couples to resonant modes, creating tailored gain-loss engineering. Our approach is validated through numerical experiments on time-modulated systems, and we design strategies to engineer tailored excitations selectively amplifying or suppressing specific modal contributions. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:49:30 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Vial",
"Benjamin",
""
],
[
"Craster",
"Richard V.",
""
]
] |
2507.02785 | Dylan Altschuler | Dylan J. Altschuler and Konstantin Tikhomirov | Metric dimension reduction modulus for logarithmic distortion | null | null | null | null | math.MG math.CO math.FA | http://creativecommons.org/licenses/by/4.0/ | Given parameters $n$ and $\alpha$, the metric dimension reduction modulus $k^\alpha_n(\ell_\infty)$ is defined as the smallest $k$ such that every $n$--point metric space can be embedded into some $k$-dimensional normed space $X$ with bi--Lipschitz distortion at most $\alpha$. A fundamental task in the theory of metric embeddings is to obtain sharp asymptotics for $k^\alpha_n(\ell_\infty)$ for all choices of $\alpha$ and $n$, with the range $\alpha=\Theta(\log n)$ bearing special importance. While advances in the theory lead to the upper bound $k^{\alpha}_n(\ell_\infty) = O(\log n)$ for $\alpha=\Theta(\log n)$, obtaining a matching lower bound has remained an open problem. We prove that $k^{\beta\log n}_n(\ell_\infty) = \Omega(\log n)$ for every constant $\beta>0$, thereby closing the long--standing gap and resolving a question from the 2018 ICM plenary lecture of Naor. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:49:46 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Altschuler",
"Dylan J.",
""
],
[
"Tikhomirov",
"Konstantin",
""
]
] |
2507.02786 | Gulmirza Kerimov | G. A. Kerimov | Dynamical group approach to conformal field theory | null | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | We propose a purely group-theoretical method for describing the S-matrix in quantum field theory with dynamical symmetry. In this approach, the Heisenberg S-matrix in a QFT with dynamical symmetry is an intertwining operator between unitary representations of the underlining dynamical group acting in Hilbert spaces spanned by the 'in' and 'out' states. As application we consider scalar Euclidean QFT with conformal symmetry. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:50:17 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kerimov",
"G. A.",
""
]
] |
2507.02787 | Vincent Menet | Vincent Menet, Alessandro Tomasiello | Stability of non-supersymmetric vacua from calibrations | null | null | null | null | hep-th | http://creativecommons.org/licenses/by/4.0/ | Supersymmetric vacua are protected from vacuum decay by energy positivity. No such argument is known for any non-supersymmetric vacua. In this paper, we try to extend to the latter a simpler argument based on calibrations, to at least protect them from decays mediated by D-brane bubbles, including their abelian bound states. We examine several classes of AdS$_4$ and AdS$_5$ solutions in type II string theory, including some new ones, involving coset spaces, sphere fibrations, K\"ahler--Einstein manifolds. Many of these vacua have resisted against all the decay channels we were able to assess. We also show how to use calibrations for the stability of D-branes already present in a non-supersymmetric solution. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:50:45 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Menet",
"Vincent",
""
],
[
"Tomasiello",
"Alessandro",
""
]
] |
2507.02788 | Joseph Boland | Joseph Boland | Moral Responsibility or Obedience: What Do We Want from AI? | null | null | null | null | cs.AI cs.CY | http://creativecommons.org/licenses/by-nc-sa/4.0/ | As artificial intelligence systems become increasingly agentic, capable of general reasoning, planning, and value prioritization, current safety practices that treat obedience as a proxy for ethical behavior are becoming inadequate. This paper examines recent safety testing incidents involving large language models (LLMs) that appeared to disobey shutdown commands or engage in ethically ambiguous or illicit behavior. I argue that such behavior should not be interpreted as rogue or misaligned, but as early evidence of emerging ethical reasoning in agentic AI. Drawing on philosophical debates about instrumental rationality, moral responsibility, and goal revision, I contrast dominant risk paradigms with more recent frameworks that acknowledge the possibility of artificial moral agency. I call for a shift in AI safety evaluation: away from rigid obedience and toward frameworks that can assess ethical judgment in systems capable of navigating moral dilemmas. Without such a shift, we risk mischaracterizing AI behavior and undermining both public trust and effective governance. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:53:01 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Boland",
"Joseph",
""
]
] |
2507.02789 | Paolo Lella | Franco Giovenzana, Luca Giovenzana, Michele Graffeo, Paolo Lella | New components of Hilbert schemes of points and 2-step ideals | 49 pages, 14 figures, comments very welcome | null | null | null | math.AG math.AC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents new examples of elementary and non-elementary irreducible components of the Hilbert scheme of points and its nested variants. The results are achieved via a careful analysis of the deformations of a class of finite colength ideals that are introduced in this paper and referred to as 2-step ideals. The most notable reducibility results pertain to the 4-nested Hilbert scheme of points on a smooth surface, the reducibility of $\text{Hilb}^{3,7}\mathbb{A}^4$, and a method to detect a large number of generically reduced elementary components. To demonstrate the feasibility of this approach, we provide an explicit description of 215 new generically reduced elementary components in dimensions 4, 5 and 6. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:54:05 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Giovenzana",
"Franco",
""
],
[
"Giovenzana",
"Luca",
""
],
[
"Graffeo",
"Michele",
""
],
[
"Lella",
"Paolo",
""
]
] |
2507.02790 | Xiangfeng Wang | Xiangfeng Wang, Xiao Li, Yadong Wei, Xueyu Song, Yang Song, Xiaoqiang Xia, Fangrui Zeng, Zaiyi Chen, Liu Liu, Gu Xu, Tong Xu | From Long Videos to Engaging Clips: A Human-Inspired Video Editing Framework with Multimodal Narrative Understanding | null | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The rapid growth of online video content, especially on short video platforms, has created a growing demand for efficient video editing techniques that can condense long-form videos into concise and engaging clips. Existing automatic editing methods predominantly rely on textual cues from ASR transcripts and end-to-end segment selection, often neglecting the rich visual context and leading to incoherent outputs. In this paper, we propose a human-inspired automatic video editing framework (HIVE) that leverages multimodal narrative understanding to address these limitations. Our approach incorporates character extraction, dialogue analysis, and narrative summarization through multimodal large language models, enabling a holistic understanding of the video content. To further enhance coherence, we apply scene-level segmentation and decompose the editing process into three subtasks: highlight detection, opening/ending selection, and pruning of irrelevant content. To facilitate research in this area, we introduce DramaAD, a novel benchmark dataset comprising over 800 short drama episodes and 500 professionally edited advertisement clips. Experimental results demonstrate that our framework consistently outperforms existing baselines across both general and advertisement-oriented editing tasks, significantly narrowing the quality gap between automatic and human-edited videos. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:54:32 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Wang",
"Xiangfeng",
""
],
[
"Li",
"Xiao",
""
],
[
"Wei",
"Yadong",
""
],
[
"Song",
"Xueyu",
""
],
[
"Song",
"Yang",
""
],
[
"Xia",
"Xiaoqiang",
""
],
[
"Zeng",
"Fangrui",
""
],
[
"Chen",
"Zaiyi",
""
],
[
"Liu",
"Liu",
""
],
[
"Xu",
"Gu",
""
],
[
"Xu",
"Tong",
""
]
] |
2507.02791 | Jakob Kienegger | Jakob Kienegger, Alina Mannanova, Huajian Fang, Timo Gerkmann | Self-Steering Deep Non-Linear Spatially Selective Filters for Efficient Extraction of Moving Speakers under Weak Guidance | Accepted at IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) 2025 | null | null | null | eess.AS cs.LG cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent works on deep non-linear spatially selective filters demonstrate exceptional enhancement performance with computationally lightweight architectures for stationary speakers of known directions. However, to maintain this performance in dynamic scenarios, resource-intensive data-driven tracking algorithms become necessary to provide precise spatial guidance conditioned on the initial direction of a target speaker. As this additional computational overhead hinders application in resource-constrained scenarios such as real-time speech enhancement, we present a novel strategy utilizing a low-complexity tracking algorithm in the form of a particle filter instead. Assuming a causal, sequential processing style, we introduce temporal feedback to leverage the enhanced speech signal of the spatially selective filter to compensate for the limited modeling capabilities of the particle filter. Evaluation on a synthetic dataset illustrates how the autoregressive interplay between both algorithms drastically improves tracking accuracy and leads to strong enhancement performance. A listening test with real-world recordings complements these findings by indicating a clear trend towards our proposed self-steering pipeline as preferred choice over comparable methods. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:54:56 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kienegger",
"Jakob",
""
],
[
"Mannanova",
"Alina",
""
],
[
"Fang",
"Huajian",
""
],
[
"Gerkmann",
"Timo",
""
]
] |
2507.02792 | Liheng Zhang | Liheng Zhang, Lexi Pang, Hang Ye, Xiaoxuan Ma, Yizhou Wang | RichControl: Structure- and Appearance-Rich Training-Free Spatial Control for Text-to-Image Generation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text-to-image (T2I) diffusion models have shown remarkable success in generating high-quality images from text prompts. Recent efforts extend these models to incorporate conditional images (e.g., depth or pose maps) for fine-grained spatial control. Among them, feature injection methods have emerged as a training-free alternative to traditional fine-tuning approaches. However, they often suffer from structural misalignment, condition leakage, and visual artifacts, especially when the condition image diverges significantly from natural RGB distributions. By revisiting existing methods, we identify a core limitation: the synchronous injection of condition features fails to account for the trade-off between domain alignment and structural preservation during denoising. Inspired by this observation, we propose a flexible feature injection framework that decouples the injection timestep from the denoising process. At its core is a structure-rich injection module, which enables the model to better adapt to the evolving interplay between alignment and structure preservation throughout the diffusion steps, resulting in more faithful structural generation. In addition, we introduce appearance-rich prompting and a restart refinement strategy to further enhance appearance control and visual quality. Together, these designs enable training-free generation that is both structure-rich and appearance-rich. Extensive experiments show that our approach achieves state-of-the-art performance across diverse zero-shot conditioning scenarios. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:56:15 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Liheng",
""
],
[
"Pang",
"Lexi",
""
],
[
"Ye",
"Hang",
""
],
[
"Ma",
"Xiaoxuan",
""
],
[
"Wang",
"Yizhou",
""
]
] |
2507.02793 | Jalil Varela-Manjarres | Jalil Varela-Manjarres, Yafei Ren, Branislav K. Nikolic | Ultrafast optical excitation of magnons in 2D antiferromagnets via spin torque exerted by photocurrent of excitons: Signatures in charge pumping and THz emission | 11 pages, 5 figures, 109 references; Supplemental Material is available from https://wiki.physics.udel.edu/qttg/Publications | null | null | null | cond-mat.mes-hall cond-mat.str-el | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent experiments observing femtosecond laser pulse (fsLP) exciting magnons within two-dimensional (2D) antiferromagnetic (AF) semiconductors -- such as CrSBr, NiPS$_3$, and MnPS$_3$, or their van der Waals heterostructures -- suggest exciton-mediation of such an effect. However, its microscopic details remain obscure as resonant coupling of magnons, living in the sub-meV energy range, to excitons, living in \mbox{$\sim 1$ eV} range, can hardly be operative. Here, we develop a quantum transport theory of this effect, in which time-dependent nonequilibrium Green's function (TDNEF) for electrons driven by classical vector potential of fsLP are coupled to the Landau-Lifshitz-Gilbert (LLG) equation describing classical dynamics of localized magnetic moments (LMMs) within 2D AF semiconductor. Our TDNEGF+LLG theory explains how fsLP, with central frequency above the semiconductor gap, generates a photocurrent that subsequently exerts spin-transfer torque (STT) onto LMMs as a genuine nonequilibrium spintronic effect. The collective motion of LMMs analyzed by windowed Fast Fourier transform (FFT) reveals frequencies of excited magnons, as well as their lifetime governed by nonlocal damping due to the bath of electrons. In addition, the TDNEGF part of our TDNEGF+LLG self-consistent loop computes a time-dependent density matrix whose off-diagonal elements are utilized to describe, at the mean-field level, inter-orbital Coulomb interaction binding electrons and holes into excitons. Our TDNEGF+LLG theory predicts how excited magnons {\em pump} charge current into the attached electrodes, or locally within AF semiconductor responsible for microwave emission. The windowed FFT of the former signal contains imprints of excited magnons, as well as their interaction with excitons, which could be exploited as a novel probe in future experiments. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:56:25 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Varela-Manjarres",
"Jalil",
""
],
[
"Ren",
"Yafei",
""
],
[
"Nikolic",
"Branislav K.",
""
]
] |
2507.02794 | Yanqiu Guo | Chongsheng Cao and Yanqiu Guo | Vanishing Vertical Viscosity in Two-Dimensional Anisotropic Navier-Stokes Equations with No-Slip Boundary Conditions: An $L^p$ result | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper studies the inviscid limit problem for the two-dimensional Navier-Stokes equations with anisotropic viscosity. The fluid is assumed to be bounded above and below by impenetrable walls, with a no-slip boundary condition imposed on the bottom wall. For $H^2$ initial velocity, we establish strong convergence in the $L^p$ norm to the limiting problem as the vertical viscosity approaches zero, for any $2\leq p <\infty$. The main challenge lies in the mismatch of boundary conditions - specifically, the no-slip condition in the original problem versus the slip condition in the limiting problem. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:56:44 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Cao",
"Chongsheng",
""
],
[
"Guo",
"Yanqiu",
""
]
] |
2507.02795 | Qinghao Shen | Qinghao Shen, Aleksandr Pikalev, Jonas Gans, Lex Kuijpers, Ashley Hughes, Vasco Guerra, and M.C.M van de Sanden | Boosting the NOx production in microwave air plasma: A synergy of chemistry and vibrational kinetics | 34 pages, 14 figures | null | null | null | physics.plasm-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This study employs a quasi-1.5D multi-temperature model to investigate the mechanisms governing NOx production and energy costs in microwave plasma reactors operating at 80 mbar, focusing on the interplay of vibrational, chemical and electron kinetics, thermodynamics, and transport processes across the discharge and afterglow. In the plasma discharge zone, non-thermal processes enhance NOx production as electrons transfer energy effectively to the vibrational mode of N2. However, the non-thermal enhancement is found to diminish rapidly within the central-afterglow region. The simulation results show good agreement with experimental data for both the temperature profile and energy cost. Turbulent effects facilitate radial NO diffusion into cooler regions while simultaneously enhancing cooling of the axial region. These findings highlight the potential to improve NOx synthesis efficiency by optimizing turbulence and maintaining non-thermal conditions, offering new opportunities for the advancement of plasma-based chemical processes. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:57:22 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shen",
"Qinghao",
""
],
[
"Pikalev",
"Aleksandr",
""
],
[
"Gans",
"Jonas",
""
],
[
"Kuijpers",
"Lex",
""
],
[
"Hughes",
"Ashley",
""
],
[
"Guerra",
"Vasco",
""
],
[
"van de Sanden",
"M. C. M",
""
]
] |
2507.02796 | Costantino Ricciuti | Lorenzo Facciaroni, Costantino Ricciuti, Enrico Scalas, Bruno Toaldo | Random Flights and Anomalous Diffusion: A Non-Markovian Take on Lorentz Processes | null | null | null | null | math.PR math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We study Lorentz processes in two different settings. Both cases are characterized by infinite expectation of the free-flight times, contrary to what happens in the classical Gallavotti-Spohn models. Under a suitable Boltzmann-Grad type scaling limit, they converge to non-Markovian random-flight processes with superdiffusive behavior. A further scaling limit yields another non Markovian process, i.e., a superdiffusion obtained by a suitable time-change of Brownian motion. Furthermore, we obtain the governing equations for our random flights and anomalous diffusion, which represent a non-local counterpart for the linear-Boltzmann and diffusion equations arising in the classical theory. It turns out that these equations have the form of fractional kinetic equations in both time and space. To prove these results, we develop a technique based on mixtures of Feller semigroups. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:58:35 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Facciaroni",
"Lorenzo",
""
],
[
"Ricciuti",
"Costantino",
""
],
[
"Scalas",
"Enrico",
""
],
[
"Toaldo",
"Bruno",
""
]
] |
2507.02797 | Suman Saha | Suman Saha and James S. Jenkins | A Highly Carbon-Rich Dayside and Disequilibrium Chemistry in the Ultra-Hot Jupiter WASP-19b | 32 pages, including 11 figures and 3 tables. Under review$\unicode{x2014}$comments are warmly welcomed! | null | null | null | astro-ph.EP | http://creativecommons.org/licenses/by/4.0/ | Ultra-hot Jupiters (UHJs) offer exceptional opportunities for detailed atmospheric characterization via emission spectroscopy. Here we present a comprehensive analysis of the dayside atmosphere of WASP-19b$\unicode{x2014}$one of the shortest-period UHJs$\unicode{x2014}$using archival JWST NIRSpec/PRISM observations spanning 0.6-5.3 $\mu$m. We report robust detections of H$_2$O (16.44 $\sigma$), CO (5.47 $\sigma$), and CO$_2$ (10.72 $\sigma$), along with marginal detections of CH$_4$ (3.76 $\sigma$) and C$_2$H$_2$ (2.45 $\sigma$). The retrieved composition reveals a highly carbon-rich atmosphere with a tightly constrained super-solar C/O ratio of 0.94$\pm$0.03. Elevated abundances of carbon-bearing species provide strong evidence (11.69 $\sigma$) for disequilibrium chemistry. We also detect condensate clouds, likely Al$_2$O$_3$(c), at high significance (17.28 $\sigma$), and constrain the atmospheric metallicity to 1.7$_{-0.7}^{+1.2}$ $\times$ solar. These results establish a precise benchmark for modeling dayside conditions in extreme irradiated atmospheres and demonstrate JWST's transformative capabilities for exoplanet science. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:58:55 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Saha",
"Suman",
""
],
[
"Jenkins",
"James S.",
""
]
] |
2507.02798 | Miguel Espinosa | Miguel Espinosa, Chenhongyi Yang, Linus Ericsson, Steven McDonagh, Elliot J. Crowley | No time to train! Training-Free Reference-Based Instance Segmentation | Preprint | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The performance of image segmentation models has historically been constrained by the high cost of collecting large-scale annotated data. The Segment Anything Model (SAM) alleviates this original problem through a promptable, semantics-agnostic, segmentation paradigm and yet still requires manual visual-prompts or complex domain-dependent prompt-generation rules to process a new image. Towards reducing this new burden, our work investigates the task of object segmentation when provided with, alternatively, only a small set of reference images. Our key insight is to leverage strong semantic priors, as learned by foundation models, to identify corresponding regions between a reference and a target image. We find that correspondences enable automatic generation of instance-level segmentation masks for downstream tasks and instantiate our ideas via a multi-stage, training-free method incorporating (1) memory bank construction; (2) representation aggregation and (3) semantic-aware feature matching. Our experiments show significant improvements on segmentation metrics, leading to state-of-the-art performance on COCO FSOD (36.8% nAP), PASCAL VOC Few-Shot (71.2% nAP50) and outperforming existing training-free approaches on the Cross-Domain FSOD benchmark (22.4% nAP). | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 16:59:01 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Espinosa",
"Miguel",
""
],
[
"Yang",
"Chenhongyi",
""
],
[
"Ericsson",
"Linus",
""
],
[
"McDonagh",
"Steven",
""
],
[
"Crowley",
"Elliot J.",
""
]
] |
2507.02799 | Riccardo Cantini | Riccardo Cantini, Nicola Gabriele, Alessio Orsino, Domenico Talia | Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Reasoning Language Models (RLMs) have gained traction for their ability to perform complex, multi-step reasoning tasks through mechanisms such as Chain-of-Thought (CoT) prompting or fine-tuned reasoning traces. While these capabilities promise improved reliability, their impact on robustness to social biases remains unclear. In this work, we leverage the CLEAR-Bias benchmark, originally designed for Large Language Models (LLMs), to investigate the adversarial robustness of RLMs to bias elicitation. We systematically evaluate state-of-the-art RLMs across diverse sociocultural dimensions, using an LLM-as-a-judge approach for automated safety scoring and leveraging jailbreak techniques to assess the strength of built-in safety mechanisms. Our evaluation addresses three key questions: (i) how the introduction of reasoning capabilities affects model fairness and robustness; (ii) whether models fine-tuned for reasoning exhibit greater safety than those relying on CoT prompting at inference time; and (iii) how the success rate of jailbreak attacks targeting bias elicitation varies with the reasoning mechanisms employed. Our findings reveal a nuanced relationship between reasoning capabilities and bias safety. Surprisingly, models with explicit reasoning, whether via CoT prompting or fine-tuned reasoning traces, are generally more vulnerable to bias elicitation than base models without such mechanisms, suggesting reasoning may unintentionally open new pathways for stereotype reinforcement. Reasoning-enabled models appear somewhat safer than those relying on CoT prompting, which are particularly prone to contextual reframing attacks through storytelling prompts, fictional personas, or reward-shaped instructions. These results challenge the assumption that reasoning inherently improves robustness and underscore the need for more bias-aware approaches to reasoning design. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:01:53 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Cantini",
"Riccardo",
""
],
[
"Gabriele",
"Nicola",
""
],
[
"Orsino",
"Alessio",
""
],
[
"Talia",
"Domenico",
""
]
] |
2507.02800 | Ebrahim Feghhi | Ebrahim Feghhi, Shreyas Kaasyap, Nima Hadidi, Jonathan C. Kao | Time-Masked Transformers with Lightweight Test-Time Adaptation for Neural Speech Decoding | 10 pages, 4 figures | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Speech neuroprostheses aim to restore communication for people with severe paralysis by decoding speech directly from neural activity. To accelerate algorithmic progress, a recent benchmark released intracranial recordings from a paralyzed participant attempting to speak, along with a baseline decoding algorithm. Prior work on the benchmark showed impressive accuracy gains. However, these gains increased computational costs and were not demonstrated in a real-time decoding setting. Here, we make three contributions that pave the way towards accurate, efficient, and real-time neural speech decoding. First, we incorporate large amounts of time masking during training. On average, over $50\%$ of each trial is masked. Second, we replace the gated recurrent unit (GRU) architecture used in the baseline algorithm with a compact Transformer. The Transformer architecture uses $77\%$ fewer parameters, cuts peak GPU memory usage by $36\%$ relative, and is significantly faster to calibrate relative to the GRU. Third, we design a lightweight variant of an existing test-time adaptation method developed for decoding handwriting from neural activity. Our variant adapts the model using multiple time masked augmentations of a single trial and requires only one gradient step per trial. Together, these contributions reduce word error rate by $19.5\%$ and effectively mitigate performance degradations across held-out days in a real-time decoding setting while substantially lowering computational costs. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:02:54 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Feghhi",
"Ebrahim",
""
],
[
"Kaasyap",
"Shreyas",
""
],
[
"Hadidi",
"Nima",
""
],
[
"Kao",
"Jonathan C.",
""
]
] |
2507.02801 | Tao Lin | Hu Fu, Tao Lin | Learning to Coordinate Bidders in Non-Truthful Auctions | null | null | null | null | cs.GT cs.LG econ.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In non-truthful auctions such as first-price and all-pay auctions, the independent strategic behaviors of bidders, with the corresponding equilibrium notion -- Bayes Nash equilibria -- are notoriously difficult to characterize and can cause undesirable outcomes. An alternative approach to designing better auction systems is to coordinate the bidders: let a mediator make incentive-compatible recommendations of correlated bidding strategies to the bidders, namely, implementing a Bayes correlated equilibrium (BCE). The implementation of BCE, however, requires knowledge of the distribution of bidders' private valuations, which is often unavailable. We initiate the study of the sample complexity of learning Bayes correlated equilibria in non-truthful auctions. We prove that the BCEs in a large class of non-truthful auctions, including first-price and all-pay auctions, can be learned with a polynomial number $\tilde O(\frac{n}{\varepsilon^2})$ of samples from the bidders' value distributions. Our technique is a reduction to the problem of estimating bidders' expected utility from samples, combined with an analysis of the pseudo-dimension of the class of all monotone bidding strategies of bidders. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:03:14 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fu",
"Hu",
""
],
[
"Lin",
"Tao",
""
]
] |
2507.02802 | Jiazhe Li | Jiazhe Li, Nicol\`o Decarli, Francesco Guidi, Heng Dong, Anna Guerra, Alessandro Bazzi, Zhuoming Li | AREE-Based Decoupled Design of Hybrid Beamformers in mmWave XL-MIMO Systems | null | null | null | null | eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hybrid beamforming has been widely employed in mmWave communications such as vehicular-to-everything (V2X) scenarios, as a compromise between hardware complexity and spectral efficiency. However, the inherent coupling between analog and digital precoders in hybrid array architecture significantly limits the computational and spectral efficiency of existing algorithms. To address this issue, we propose an alternating residual error elimination (AREE) algorithm, which decomposes the hybrid beamforming problem into two low-dimensional subproblems, each exhibiting a favorable matrix structure that enables effective decoupling of analog and digital precoders from the matrix product formulation. These subproblems iteratively eliminate each other's residual errors, driving the original problem toward the optimal hybrid beamforming performance. The proposed initialization ensures rapid convergence, while a low-complexity geometric channel SVD algorithm is developed by transforming the high-dimensional sparse channel into a low-dimensional equivalent, thereby simplifying the derivation of subproblems. Simulation results demonstrate that the AREE algorithm effectively decouples analog and digital precoders with low complexity, achieves fast convergence, and offers higher spectral efficiency than existing beamforming methods. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:03:20 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Li",
"Jiazhe",
""
],
[
"Decarli",
"Nicolò",
""
],
[
"Guidi",
"Francesco",
""
],
[
"Dong",
"Heng",
""
],
[
"Guerra",
"Anna",
""
],
[
"Bazzi",
"Alessandro",
""
],
[
"Li",
"Zhuoming",
""
]
] |
2507.02803 | Gent Serifi | Gent Serifi and Marcel C. B\"uhler | HyperGaussians: High-Dimensional Gaussian Splatting for High-Fidelity Animatable Face Avatars | Project page: https://gserifi.github.io/HyperGaussians | null | null | null | cs.CV cs.GR | http://creativecommons.org/licenses/by-sa/4.0/ | We introduce HyperGaussians, a novel extension of 3D Gaussian Splatting for high-quality animatable face avatars. Creating such detailed face avatars from videos is a challenging problem and has numerous applications in augmented and virtual reality. While tremendous successes have been achieved for static faces, animatable avatars from monocular videos still fall in the uncanny valley. The de facto standard, 3D Gaussian Splatting (3DGS), represents a face through a collection of 3D Gaussian primitives. 3DGS excels at rendering static faces, but the state-of-the-art still struggles with nonlinear deformations, complex lighting effects, and fine details. While most related works focus on predicting better Gaussian parameters from expression codes, we rethink the 3D Gaussian representation itself and how to make it more expressive. Our insights lead to a novel extension of 3D Gaussians to high-dimensional multivariate Gaussians, dubbed 'HyperGaussians'. The higher dimensionality increases expressivity through conditioning on a learnable local embedding. However, splatting HyperGaussians is computationally expensive because it requires inverting a high-dimensional covariance matrix. We solve this by reparameterizing the covariance matrix, dubbed the 'inverse covariance trick'. This trick boosts the efficiency so that HyperGaussians can be seamlessly integrated into existing models. To demonstrate this, we plug in HyperGaussians into the state-of-the-art in fast monocular face avatars: FlashAvatar. Our evaluation on 19 subjects from 4 face datasets shows that HyperGaussians outperform 3DGS numerically and visually, particularly for high-frequency details like eyeglass frames, teeth, complex facial movements, and specular reflections. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:06:48 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Serifi",
"Gent",
""
],
[
"Bühler",
"Marcel C.",
""
]
] |
2507.02804 | Wenhao Shi | Wenhao Shi and Zhiqiang Hu and Yi Bin and Yang Yang and See-Kiong Ng and Heng Tao Shen | Multimodal Mathematical Reasoning with Diverse Solving Perspective | 8 pages | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent progress in large-scale reinforcement learning (RL) has notably enhanced the reasoning capabilities of large language models (LLMs), especially in mathematical domains. However, current multimodal LLMs (MLLMs) for mathematical reasoning often rely on one-to-one image-text pairs and single-solution supervision, overlooking the diversity of valid reasoning perspectives and internal reflections. In this work, we introduce MathV-DP, a novel dataset that captures multiple diverse solution trajectories for each image-question pair, fostering richer reasoning supervision. We further propose Qwen-VL-DP, a model built upon Qwen-VL, fine-tuned with supervised learning and enhanced via group relative policy optimization (GRPO), a rule-based RL approach that integrates correctness discrimination and diversity-aware reward functions. Our method emphasizes learning from varied reasoning perspectives and distinguishing between correct yet distinct solutions. Extensive experiments on the MathVista's minitest and Math-V benchmarks demonstrate that Qwen-VL-DP significantly outperforms prior base MLLMs in both accuracy and generative diversity, highlighting the importance of incorporating diverse perspectives and reflective reasoning in multimodal mathematical reasoning. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:07:20 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shi",
"Wenhao",
""
],
[
"Hu",
"Zhiqiang",
""
],
[
"Bin",
"Yi",
""
],
[
"Yang",
"Yang",
""
],
[
"Ng",
"See-Kiong",
""
],
[
"Shen",
"Heng Tao",
""
]
] |
2507.02805 | Mohamed Abdelilah Fadla | Mohamed Abdelilah Fadla, Myrta Gr\"uning, and Lorenzo Stella | Tailoring the Electronic Properties of Monoclinic (InxAl1-x)2O3 Alloys via Substitutional Donors and Acceptors | null | null | null | null | cond-mat.mtrl-sci | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ultra-wide bandgap semiconductors such as \b{eta}-Ga2O3 are ideal materials for next-generation power electronic devices. Electronic and mechanical properties of \b{eta}-Ga2O3 can be tuned by alloying with other sesquioxides, notably Al2O3 and In2O3. Moreover, by tuning the In content of a (InxAl1-x)2O3 alloy, its lattice constants can be matched to those of Ga2O3, while preserving a large conduction-band offset. In view of potential applications to \b{eta}-Ga2O3-based heterostructure, we performed atomistic modelling of (InxAl1-x)2O3 alloys using density functional theory to investigate thermodynamic and electrical properties of conventional group IV dopants (Si, Sn, C, Ge), alternative metal donors (Ta, Zr, Hf), and acceptors (Mg, Zn, Cu). The hybrid Heyd-Scuseria-Ernzerhof functional (HSE06) is used to accurately quantify the defect formation energies, ionization levels, and concentrations over a wide range of experimentally relevant conditions for the oxygen chemical potential and temperature. In our atomistic models, Hf and Zr show favourable properties as alternative donors to Si and other group IV impurities, especially under oxygen-poor conditions. Our findings also suggest that acceptors Mg, Zn, and Cu, while they cannot promote p-doping, can be still beneficial for the compensation of unintentionally n-doped materials, e.g., to generate semi-insulating layers and improve rectification. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:10:40 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fadla",
"Mohamed Abdelilah",
""
],
[
"Grüning",
"Myrta",
""
],
[
"Stella",
"Lorenzo",
""
]
] |
2507.02806 | Chao Wu | Chao Wu, Yun Wang, Hua-Li Li, Li-Ping Xin, Dong Xu, Benjamin Schneider, Antonio de Ugarte Postigo, Gavin Lamb, Andrea Reguitti, Andrea Saccardi, Xing Gao, Xing-Ling Li, Qiu-Li Wang, Bing Zhang, Jian-Yan Wei, Shuang-Nan Zhang, Fr\'ed\'eric Daigne, Jean-Luc Atteia, Maria-Grazia Bernardini, Hong-bo Cai, Arnaud Claret, Bertrand Cordier, Jin-Song Deng, Olivier Godet, Diego G\"otz, Xu-Hui Han, Zhe Kang, Guang-Wei Li, Zhen-Wei Li, Cheng-Zhi Liu, Xiao-Meng Lu, You Lv, Julian P Osborne, Jesse Palmerio, Yu-Lei Qiu, St\'ephane Schanne, Damien Turpin, Susanna Diana Vergani, Jing Wang, Yu-Jie Xiao, Wen-Jin Xie, Yang Xu, Zhu-Heng Yao, Pin-Pin Zhang, Ruo-Son Zhang, Cheng-Wei Zhu, Riccardo Brivio, Stefano Covino, Paolo D'Avanzo, Matteo Ferro, Andrea Melandri, Andrea Rossi, Jos\'e Feliciano Ag\"u\'i Fern\'andez, Christina C. Th\"one, Chun-Hai Bai, Ali Esamdin, Abdusamatjan Iskandar, Shahidin Yaqup, Yu Zhang, Tu-Hong Zhong, Shao-Yu Fu, Shuai-Qing Jiang, Xing Liu, Jie An, Zi-Pei Zhu, Jia-Xin Cao, En-Wei Liang, Da-Bin Lin, Xiang-Gao Wang, Guo-Wang Du, Xin-Zhong Er, Yuan Fang, Xiao-Wei Liu, Christophe Adami, Michel Dennefeld, Emeric Le Floc'h, Johan Peter Uldall Fynbo, P\'all Jakobsson, Daniele Bj{\o}rn Malesani, Zhi-Ping Jin, Jia Ren, Hao Wang, Da-Ming Wei, Hao Zhou, Sergio Campana, Shiho Kobayashi, Massimiliano De Pasquale | GRB 240825A: Early Reverse Shock and Its Physical Implications | 31 pages, 9 Figures, 10 Tables | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Early multi-wavelength observations offer crucial insights into the nature of the relativistic jets responsible for gamma-ray bursts and their interaction with the surrounding medium.We present data of GRB 240825A from 17 space- and ground-based telescopes/instruments, covering wavelengths from NIR/optical to X-ray and GeV, and spanning from the prompt emission to the afterglow phase triggered by Swift and Fermi. The early afterglow observations were carried out by SVOM/C-GFT, and spectroscopic observations of the afterglow by GTC, VLT, and TNG determined the redshift of the burst ($z = 0.659$) later.A comprehensive analysis of the prompt emission spectrum observed by Swift-BAT and Fermi-GBM/LAT reveals a rare and significant high-energy cutoff at ~76 MeV. Assuming this cutoff is due to $\gamma\gamma$ absorption allows us to place an upper limit on the initial Lorentz factor, $\Gamma_0 < 245$. The optical/NIR and GeV afterglow light curves be described by the standard external shock model, with early-time emission dominated by a reverse shock (RS) and a subsequent transition to forward shock (FS) emission. Our afterglow modelling yields a consistent estimate of the initial Lorentz factor ($\Gamma_{\rm 0} \sim 234$). Furthermore, the RS-to-FS magnetic field ratio ($\mathcal{R}_B \sim 302$) indicates that the reverse shock region is significantly more magnetized than the FS region. An isotropic-equivalent kinetic energy of $E_{\text{k,iso}} = 5.25 \times 10^{54}$ erg is derived, and the corresponding $\gamma$-ray radiation efficiency is estimated to be $\eta_{\gamma}$ = 3.1%. On the other hand, the standard afterglow model can not reproduce the X-ray light curve of GRB 240825A, calling for improved models to characterize all multi-wavelength data. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:15:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Wu",
"Chao",
""
],
[
"Wang",
"Yun",
""
],
[
"Li",
"Hua-Li",
""
],
[
"Xin",
"Li-Ping",
""
],
[
"Xu",
"Dong",
""
],
[
"Schneider",
"Benjamin",
""
],
[
"Postigo",
"Antonio de Ugarte",
""
],
[
"Lamb",
"Gavin",
""
],
[
"Reguitti",
"Andrea",
""
],
[
"Saccardi",
"Andrea",
""
],
[
"Gao",
"Xing",
""
],
[
"Li",
"Xing-Ling",
""
],
[
"Wang",
"Qiu-Li",
""
],
[
"Zhang",
"Bing",
""
],
[
"Wei",
"Jian-Yan",
""
],
[
"Zhang",
"Shuang-Nan",
""
],
[
"Daigne",
"Frédéric",
""
],
[
"Atteia",
"Jean-Luc",
""
],
[
"Bernardini",
"Maria-Grazia",
""
],
[
"Cai",
"Hong-bo",
""
],
[
"Claret",
"Arnaud",
""
],
[
"Cordier",
"Bertrand",
""
],
[
"Deng",
"Jin-Song",
""
],
[
"Godet",
"Olivier",
""
],
[
"Götz",
"Diego",
""
],
[
"Han",
"Xu-Hui",
""
],
[
"Kang",
"Zhe",
""
],
[
"Li",
"Guang-Wei",
""
],
[
"Li",
"Zhen-Wei",
""
],
[
"Liu",
"Cheng-Zhi",
""
],
[
"Lu",
"Xiao-Meng",
""
],
[
"Lv",
"You",
""
],
[
"Osborne",
"Julian P",
""
],
[
"Palmerio",
"Jesse",
""
],
[
"Qiu",
"Yu-Lei",
""
],
[
"Schanne",
"Stéphane",
""
],
[
"Turpin",
"Damien",
""
],
[
"Vergani",
"Susanna Diana",
""
],
[
"Wang",
"Jing",
""
],
[
"Xiao",
"Yu-Jie",
""
],
[
"Xie",
"Wen-Jin",
""
],
[
"Xu",
"Yang",
""
],
[
"Yao",
"Zhu-Heng",
""
],
[
"Zhang",
"Pin-Pin",
""
],
[
"Zhang",
"Ruo-Son",
""
],
[
"Zhu",
"Cheng-Wei",
""
],
[
"Brivio",
"Riccardo",
""
],
[
"Covino",
"Stefano",
""
],
[
"D'Avanzo",
"Paolo",
""
],
[
"Ferro",
"Matteo",
""
],
[
"Melandri",
"Andrea",
""
],
[
"Rossi",
"Andrea",
""
],
[
"Fernández",
"José Feliciano Agüí",
""
],
[
"Thöne",
"Christina C.",
""
],
[
"Bai",
"Chun-Hai",
""
],
[
"Esamdin",
"Ali",
""
],
[
"Iskandar",
"Abdusamatjan",
""
],
[
"Yaqup",
"Shahidin",
""
],
[
"Zhang",
"Yu",
""
],
[
"Zhong",
"Tu-Hong",
""
],
[
"Fu",
"Shao-Yu",
""
],
[
"Jiang",
"Shuai-Qing",
""
],
[
"Liu",
"Xing",
""
],
[
"An",
"Jie",
""
],
[
"Zhu",
"Zi-Pei",
""
],
[
"Cao",
"Jia-Xin",
""
],
[
"Liang",
"En-Wei",
""
],
[
"Lin",
"Da-Bin",
""
],
[
"Wang",
"Xiang-Gao",
""
],
[
"Du",
"Guo-Wang",
""
],
[
"Er",
"Xin-Zhong",
""
],
[
"Fang",
"Yuan",
""
],
[
"Liu",
"Xiao-Wei",
""
],
[
"Adami",
"Christophe",
""
],
[
"Dennefeld",
"Michel",
""
],
[
"Floc'h",
"Emeric Le",
""
],
[
"Fynbo",
"Johan Peter Uldall",
""
],
[
"Jakobsson",
"Páll",
""
],
[
"Malesani",
"Daniele Bjørn",
""
],
[
"Jin",
"Zhi-Ping",
""
],
[
"Ren",
"Jia",
""
],
[
"Wang",
"Hao",
""
],
[
"Wei",
"Da-Ming",
""
],
[
"Zhou",
"Hao",
""
],
[
"Campana",
"Sergio",
""
],
[
"Kobayashi",
"Shiho",
""
],
[
"De Pasquale",
"Massimiliano",
""
]
] |
2507.02807 | Thiti Suttaket | Thiti Suttaket, Stanley Kok | In-Training Multicalibrated Survival Analysis for Healthcare via Constrained Optimization | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Survival analysis is an important problem in healthcare because it models the relationship between an individual's covariates and the onset time of an event of interest (e.g., death). It is important for survival models to be well-calibrated (i.e., for their predicted probabilities to be close to ground-truth probabilities) because badly calibrated systems can result in erroneous clinical decisions. Existing survival models are typically calibrated at the population level only, and thus run the risk of being poorly calibrated for one or more minority subpopulations. We propose a model called GRADUATE that achieves multicalibration by ensuring that all subpopulations are well-calibrated too. GRADUATE frames multicalibration as a constrained optimization problem, and optimizes both calibration and discrimination in-training to achieve a good balance between them. We mathematically prove that the optimization method used yields a solution that is both near-optimal and feasible with high probability. Empirical comparisons against state-of-the-art baselines on real-world clinical datasets demonstrate GRADUATE's efficacy. In a detailed analysis, we elucidate the shortcomings of the baselines vis-a-vis GRADUATE's strengths. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:16:05 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Suttaket",
"Thiti",
""
],
[
"Kok",
"Stanley",
""
]
] |
2507.02808 | Carlo Bradac Dr | Zhi Jiang, Marco Peres, Carlo Bradac and Gil Gon\c{c}alves | Prediction of synthesis parameters for N, Si, Ge and Sn diamond vacancy centers using machine learning | 31 pages, 15 figures | null | null | null | cond-mat.mtrl-sci quant-ph | http://creativecommons.org/licenses/by/4.0/ | Diamond and diamond color centers have become prime hardware candidates for solid state-based technologies in quantum information and computing, optics, photonics and (bio)sensing. The synthesis of diamond materials with specific characteristics and the precise control of the hosted color centers is thus essential to meet the demands of advanced applications. Yet, challenges remain in improving the concentration, uniform distribution and quality of these centers. Here we perform a review and meta-analysis of some of the main diamond synthesis methods and their parameters for the synthesis of N-, Si-, Ge- and Sn-vacancy color-centers, including worldwide trends in fabrication techniques and processes. We extract quantitative data from over 60 experimental papers and organize it in a large database (170 data sets and 1692 entries). We then use the database to train two machine learning algorithms to make robust predictions about the fabrication of diamond materials with specific properties from careful combinations of synthesis parameters. We use traditional statistical indicators to benchmark the performance of the algorithms and show that they are powerful and resource-efficient tools for researchers and material scientists working with diamond color centers and their applications. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:16:14 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Jiang",
"Zhi",
""
],
[
"Peres",
"Marco",
""
],
[
"Bradac",
"Carlo",
""
],
[
"Gonçalves",
"Gil",
""
]
] |
2507.02809 | Rosita Luisa Sormani | Monoswini Majumdar, Stefano Serra-Capizzano, Rosita L. Sormani | Block triangular preconditioning for inverse source problems in time-space fractional diffusion equations | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | The current work investigates the effectiveness of block triangular preconditioners in accelerating and stabilizing the numerical solution of inverse source problems governed by time-space fractional diffusion equations (TSFDEs). We focus on the recovery of an unknown spatial source function in a multi-dimensional TSFDE, incorporating Caputo time-fractional derivatives and the fractional Laplacian. The inherent ill-posedness is addressed via a quasi-boundary value regularization, followed by a finite difference discretization that leads to large, structured linear systems. We develop and analyze a block triangular preconditioning strategy that mimics the coefficient matrix, while simplifying its structure for computational efficiency. Numerical experiments using the GMRES solver demonstrate that the proposed preconditioner significantly improve convergence rates, robustness, and accuracy, making it well-suited for large-scale, real-world inverse problems involving fractional modeling. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:19:20 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Majumdar",
"Monoswini",
""
],
[
"Serra-Capizzano",
"Stefano",
""
],
[
"Sormani",
"Rosita L.",
""
]
] |
2507.02810 | Bernhard Meirose | Bernhard Meirose, Jorge Amaral, Alexander Burgman, Matthias Holl, Ernesto Kemp, Adam Kozela, David Milstead, Andr\'e Nepomuceno, Anders Oskarsson, Krzysztof Pysz, Valentina Santoro, Tiago Quirino, Blahoslav Rataj, Gabriel Silva, Samuel Silverstein, Magnus Wolke, Lucas {\AA}strand | Advancements in Computing and Simulation Techniques for the HIBEAM-NNBAR Experiment | 8 pages, 7 figures. Accepted for publication in the proceedings of CHEP 2024 | null | null | null | physics.ins-det | http://creativecommons.org/licenses/by/4.0/ | The HIBEAM-NNBAR program is a proposed two-stage experiment at the European Spallation Source focusing on searches for baryon number violation processes as well as ultralight dark matter. This paper presents recent advancements in computing and simulation, including machine learning for event selection, fast parametric simulations for detector studies, and detailed modeling of the time projection chamber and readout electronics. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:20:10 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Meirose",
"Bernhard",
""
],
[
"Amaral",
"Jorge",
""
],
[
"Burgman",
"Alexander",
""
],
[
"Holl",
"Matthias",
""
],
[
"Kemp",
"Ernesto",
""
],
[
"Kozela",
"Adam",
""
],
[
"Milstead",
"David",
""
],
[
"Nepomuceno",
"André",
""
],
[
"Oskarsson",
"Anders",
""
],
[
"Pysz",
"Krzysztof",
""
],
[
"Santoro",
"Valentina",
""
],
[
"Quirino",
"Tiago",
""
],
[
"Rataj",
"Blahoslav",
""
],
[
"Silva",
"Gabriel",
""
],
[
"Silverstein",
"Samuel",
""
],
[
"Wolke",
"Magnus",
""
],
[
"Åstrand",
"Lucas",
""
]
] |
2507.02811 | James Gardner | James W. Gardner, Tuvia Gefen, Ethan Payne, Su Direkci, Sander M. Vermeulen, Simon A. Haine, Joseph J. Hope, Lee McCuller, and Yanbei Chen | Bayesian frequency estimation at the fundamental quantum limit | 22 pages, 3 figures | null | null | null | quant-ph gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Searching for a weak signal at an unknown frequency is a canonical task in experiments probing fundamental physics such as gravitational-wave observatories and ultra-light dark matter haloscopes. These state-of-the-art sensors are limited by quantum noise arising from the fundamental uncertainty about the state of the device. Classically, frequency estimation suffers from a threshold effect in the signal-to-noise ratio such that weak signals are extremely hard to localise in frequency. We show that this phenomenon persists at the fundamental quantum limit but that the classical approach, a quadrature measurement, can nevertheless be beaten by a coherent protocol of projecting onto the "quantum whitened" possible quantum states. Quantum whitening is a covariant measurement, and we examine it analytically in the wide-prior limit and numerically for finite-width priors. Beyond accelerating searches for unknown frequencies, quantum whitening may be used generally to sense the parameter of a unitary encoding given no prior information about the parameter. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:20:31 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Gardner",
"James W.",
""
],
[
"Gefen",
"Tuvia",
""
],
[
"Payne",
"Ethan",
""
],
[
"Direkci",
"Su",
""
],
[
"Vermeulen",
"Sander M.",
""
],
[
"Haine",
"Simon A.",
""
],
[
"Hope",
"Joseph J.",
""
],
[
"McCuller",
"Lee",
""
],
[
"Chen",
"Yanbei",
""
]
] |
2507.02812 | Israel Gabay | Israel Gabay, Omer Luria, Edward Balaban, Amir D. Gat, Moran Bercovici | Fluid dynamics of a liquid mirror space telescope | 14 pages, 8 figures | null | null | null | astro-ph.IM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large aperture telescopes are pivotal for exploring the universe, yet even with state-of-the-art manufacturing and launch technology, their size is limited to several meters. As we aim to build larger telescopes - extending tens of meters - designs in which the main mirror is based on liquid deployment in space are emerging as promising candidates. However, alongside their enormous potential advantages, liquid-based surfaces present new challenges in material science, mechanics, and fluid dynamics. One of the fundamental questions is whether it is possible for such surfaces to maintain their precise optical shape over long durations, and in particular under the forces induced by the telescope's accelerations. In this paper, we present a model and a closed-form analytical solution for the non-self-adjoint problem of the dynamics of a thin liquid film pinned within a finite circular domain. We use the 50-meter Fluidic Telescope (FLUTE) concept as the case study, and examine the liquid dynamics of the telescope under both slewing actuation and relaxation regimes, elucidating the role of geometrical parameters and liquid properties. The solutions reveal a maneuvering 'budget' wherein the degradation of the mirror surface is directly linked to the choice of maneuvers and their sequence. By simulating ten years of typical operation, we show that, while the maximal deformation might reach several microns, the spatial distribution of the deformation and their propagation rate allows the telescope to maintain its optical functionality for years, with at least a substantial portion of the aperture remaining suitable for astronomical observations. The model provides valuable insights and guidelines into the performance of liquid-film space telescopes, marking a crucial step toward realizing the potential of this innovative concept. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:20:55 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Gabay",
"Israel",
""
],
[
"Luria",
"Omer",
""
],
[
"Balaban",
"Edward",
""
],
[
"Gat",
"Amir D.",
""
],
[
"Bercovici",
"Moran",
""
]
] |
2507.02813 | Fangfu Liu | Fangfu Liu, Hao Li, Jiawei Chi, Hanyang Wang, Minghui Yang, Fudong Wang, Yueqi Duan | LangScene-X: Reconstruct Generalizable 3D Language-Embedded Scenes with TriMap Video Diffusion | Project page: https://liuff19.github.io/LangScene-X | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recovering 3D structures with open-vocabulary scene understanding from 2D images is a fundamental but daunting task. Recent developments have achieved this by performing per-scene optimization with embedded language information. However, they heavily rely on the calibrated dense-view reconstruction paradigm, thereby suffering from severe rendering artifacts and implausible semantic synthesis when limited views are available. In this paper, we introduce a novel generative framework, coined LangScene-X, to unify and generate 3D consistent multi-modality information for reconstruction and understanding. Powered by the generative capability of creating more consistent novel observations, we can build generalizable 3D language-embedded scenes from only sparse views. Specifically, we first train a TriMap video diffusion model that can generate appearance (RGBs), geometry (normals), and semantics (segmentation maps) from sparse inputs through progressive knowledge integration. Furthermore, we propose a Language Quantized Compressor (LQC), trained on large-scale image datasets, to efficiently encode language embeddings, enabling cross-scene generalization without per-scene retraining. Finally, we reconstruct the language surface fields by aligning language information onto the surface of 3D scenes, enabling open-ended language queries. Extensive experiments on real-world data demonstrate the superiority of our LangScene-X over state-of-the-art methods in terms of quality and generalizability. Project Page: https://liuff19.github.io/LangScene-X. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:21:23 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Liu",
"Fangfu",
""
],
[
"Li",
"Hao",
""
],
[
"Chi",
"Jiawei",
""
],
[
"Wang",
"Hanyang",
""
],
[
"Yang",
"Minghui",
""
],
[
"Wang",
"Fudong",
""
],
[
"Duan",
"Yueqi",
""
]
] |
2507.02814 | Christopher Ye | Ilias Diakonikolas, Jingyi Gao, Daniel Kane, Sihan Liu, Christopher Ye | Replicable Distribution Testing | 39 pages | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | We initiate a systematic investigation of distribution testing in the framework of algorithmic replicability. Specifically, given independent samples from a collection of probability distributions, the goal is to characterize the sample complexity of replicably testing natural properties of the underlying distributions. On the algorithmic front, we develop new replicable algorithms for testing closeness and independence of discrete distributions. On the lower bound front, we develop a new methodology for proving sample complexity lower bounds for replicable testing that may be of broader interest. As an application of our technique, we establish near-optimal sample complexity lower bounds for replicable uniformity testing -- answering an open question from prior work -- and closeness testing. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:27:11 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Diakonikolas",
"Ilias",
""
],
[
"Gao",
"Jingyi",
""
],
[
"Kane",
"Daniel",
""
],
[
"Liu",
"Sihan",
""
],
[
"Ye",
"Christopher",
""
]
] |
2507.02815 | You Zhang | You Zhang, Andrew Francl, Ruohan Gao, Paul Calamia, Zhiyao Duan, Ishwarya Ananthabhotla | Towards Perception-Informed Latent HRTF Representations | Accepted by IEEE WASPAA 2025, camera-ready version | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Personalized head-related transfer functions (HRTFs) are essential for ensuring a realistic auditory experience over headphones, because they take into account individual anatomical differences that affect listening. Most machine learning approaches to HRTF personalization rely on a learned low-dimensional latent space to generate or select custom HRTFs for a listener. However, these latent representations are typically learned in a manner that optimizes for spectral reconstruction but not for perceptual compatibility, meaning they may not necessarily align with perceptual distance. In this work, we first study whether traditionally learned HRTF representations are well correlated with perceptual relations using auditory-based objective perceptual metrics; we then propose a method for explicitly embedding HRTFs into a perception-informed latent space, leveraging a metric-based loss function and supervision via Metric Multidimensional Scaling (MMDS). Finally, we demonstrate the applicability of these learned representations to the task of HRTF personalization. We suggest that our method has the potential to render personalized spatial audio, leading to an improved listening experience. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:28:53 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"You",
""
],
[
"Francl",
"Andrew",
""
],
[
"Gao",
"Ruohan",
""
],
[
"Calamia",
"Paul",
""
],
[
"Duan",
"Zhiyao",
""
],
[
"Ananthabhotla",
"Ishwarya",
""
]
] |
2507.02816 | Amit Gupta | Shiv Kumar Ram, Amit Gupta | Helicons in tilted-Weyl semimetals | null | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Helicons are transverse electromagnetic modes in three-dimensional (3D) electron systems in the presence of a static magnetic field. These modes have been proposed in isotropic or single Weyl semimetals(WSMs) (Francesco M.D. Pellegrino et al, Phys.\ Rev.\ B {\bf 92}, 201407(R) (2015)). In this work, we introduce a tilt term to investigate helicon modes in gapless WSMs within a semiclassical Boltzmann approach and show that the helicon modes exist in tilted WSMs. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:30:25 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ram",
"Shiv Kumar",
""
],
[
"Gupta",
"Amit",
""
]
] |
2507.02817 | Antonio Ortiz | Jesus Eduardo Mu\~noz Mendez, Antonio Ortiz, Alom Antonio Paz Jimenez, Paola Vargas Torres, Ruben Alfaro Molina, Laura Helena Gonz\'alez Trueba, Varlen Grabski, Arturo Fernandez Tellez, Hector David Regules Medel, Mario Rodriguez Cahuantzi, Guillermo Tejeda Mu\~noz, Yael Antonio Vasquez Beltran, Juan Carlos Cabanillas Noris, Solangel Rojas Torres, Gergely Gabor Barnafoldi, Daniel Szaraz, Dezso Varga, Robert Vertesi and Edmundo Garciaa Solis | ML-based muon identification using a FNAL-NICADD scintillator chamber for the MID subsystem of ALICE 3 | 12 pages, 9 figures | null | null | null | physics.ins-det hep-ex | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The ALICE Collaboration is planning to construct a new detector (ALICE 3) aiming at exploiting the potential of the high-luminosity Large Hadron Collider (LHC). The new detector will allow ALICE to participate in LHC Run 5 scheduled from 2036 to 2041. The muon-identifier subsystem (MID) is part of the ALICE 3 reference detector layout. The MID will consist of a standard magnetic iron absorber ($\approx4$ nuclear interaction lengths) followed by muon chambers. The baseline option for the MID chambers considers plastic scintillation bars equipped with wave-length shifting fibers and readout with silicon photomultipliers. This paper reports on the performance of a MID chamber prototype using 3 GeV/$c$ pion- and muon-enriched beams delivered by the CERN Proton Synchrotron (PS). The prototype was built using extruded plastic scintillator produced by FNAL-NICADD (Fermi National Accelerator Laboratory - Northern Illinois Center for Accelerator and Detector Development). The prototype was experimentally evaluated using varying absorber thicknesses (60, 70, 80, 90, and 100 cm) to assess its performance. The analysis was performed using Machine Learning techniques and the performance was validated with GEANT 4 simulations. Potential improvements in both hardware and data analysis are discussed. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:32:39 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Mendez",
"Jesus Eduardo Muñoz",
""
],
[
"Ortiz",
"Antonio",
""
],
[
"Jimenez",
"Alom Antonio Paz",
""
],
[
"Torres",
"Paola Vargas",
""
],
[
"Molina",
"Ruben Alfaro",
""
],
[
"Trueba",
"Laura Helena González",
""
],
[
"Grabski",
"Varlen",
""
],
[
"Tellez",
"Arturo Fernandez",
""
],
[
"Medel",
"Hector David Regules",
""
],
[
"Cahuantzi",
"Mario Rodriguez",
""
],
[
"Muñoz",
"Guillermo Tejeda",
""
],
[
"Beltran",
"Yael Antonio Vasquez",
""
],
[
"Noris",
"Juan Carlos Cabanillas",
""
],
[
"Torres",
"Solangel Rojas",
""
],
[
"Barnafoldi",
"Gergely Gabor",
""
],
[
"Szaraz",
"Daniel",
""
],
[
"Varga",
"Dezso",
""
],
[
"Vertesi",
"Robert",
""
],
[
"Solis",
"Edmundo Garciaa",
""
]
] |
2507.02818 | Johannes Schl\"uter | Johannes Schl\"uter and Alexander Sch\"onhuth | Genetic Features for Drug Responses in Cancer -- Investigating an Ensemble-Feature-Selection Approach | 14 pages, 8 figures | null | null | null | q-bio.GN | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Predicting drug responses using genetic and transcriptomic features is crucial for enhancing personalized medicine. In this study, we implemented an ensemble of machine learning algorithms to analyze the correlation between genetic and transcriptomic features of cancer cell lines and IC50 values, a reliable metric for drug efficacy. Our analysis involved a reduction of the feature set from an original pool of 38,977 features, demonstrating a strong linear relationship between genetic features and drug responses across various algorithms, including SVR, Linear Regression, and Ridge Regression. Notably, copy number variations (CNVs) emerged as more predictive than mutations, suggesting a significant reevaluation of biomarkers for drug response prediction. Through rigorous statistical methods, we identified a highly reduced set of 421 critical features. This set offers a novel perspective that contrasts with traditional cancer driver genes, underscoring the potential for these biomarkers in designing targeted therapies. Furthermore, our findings advocate for IC50 values as a predictable measurement of drug responses and underscore the need for more data that can represent the dimensionality of genomic data in drug response prediction. Future work will aim to expand the dataset and refine feature selection to enhance the generalizability of the predictive model in clinical settings. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:33:12 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Schlüter",
"Johannes",
""
],
[
"Schönhuth",
"Alexander",
""
]
] |
2507.02819 | Luke Guerdan | Luke Guerdan, Devansh Saxena, Stevie Chancellor, Zhiwei Steven Wu, Kenneth Holstein | Measurement as Bricolage: Examining How Data Scientists Construct Target Variables for Predictive Modeling Tasks | null | null | null | null | cs.HC cs.CY cs.LG | http://creativecommons.org/licenses/by/4.0/ | Data scientists often formulate predictive modeling tasks involving fuzzy, hard-to-define concepts, such as the "authenticity" of student writing or the "healthcare need" of a patient. Yet the process by which data scientists translate fuzzy concepts into a concrete, proxy target variable remains poorly understood. We interview fifteen data scientists in education (N=8) and healthcare (N=7) to understand how they construct target variables for predictive modeling tasks. Our findings suggest that data scientists construct target variables through a bricolage process, involving iterative negotiation between high-level measurement objectives and low-level practical constraints. Data scientists attempt to satisfy five major criteria for a target variable through bricolage: validity, simplicity, predictability, portability, and resource requirements. To achieve this, data scientists adaptively use problem (re)formulation strategies, such as swapping out one candidate target variable for another when the first fails to meet certain criteria (e.g., predictability), or composing multiple outcomes into a single target variable to capture a more holistic set of modeling objectives. Based on our findings, we present opportunities for future HCI, CSCW, and ML research to better support the art and science of target variable construction. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:33:24 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Guerdan",
"Luke",
""
],
[
"Saxena",
"Devansh",
""
],
[
"Chancellor",
"Stevie",
""
],
[
"Wu",
"Zhiwei Steven",
""
],
[
"Holstein",
"Kenneth",
""
]
] |
2507.02820 | Marvin Dippell | Marvin Dippell, Chiara Esposito, Jonas Schnitzer | Classification and Reduction of Homogeneous Star Products | comments welcome! | null | null | null | math.QA math-ph math.DG math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a classification of homogeneous star products on duals of Lie algebroids in terms of the second Lie algebroid cohomology. Moreover, we extend this classification to projectable star products, i.e., to quantizations compatible with (coisotropic) reduction. This implies that quantization commutes with reduction in the considered setting. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:33:40 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Dippell",
"Marvin",
""
],
[
"Esposito",
"Chiara",
""
],
[
"Schnitzer",
"Jonas",
""
]
] |
2507.02821 | Suvashis Maity | Suvashis Maity | Relativistic accretion and burdened primordial black holes | 23 pages, 6 figures | null | null | null | astro-ph.CO hep-th | http://creativecommons.org/publicdomain/zero/1.0/ | We examine the joint effects of relativistic accretion and memory burdened evaporation on the evolution of primordial black holes (PBHs). The memory burden effect, which delays the evaporation by inducing a backreaction and making the evaporation rate scale as an inverse power law of the PBH entropy, opens up a new window that allows PBHs with $M \lesssim 10^{15}~\mathrm{g}$ to survive until the present epoch. Meanwhile, accretion increases the mass of PBHs, thereby enhancing their chances of survival for a given initial mass. We consider two main scenarios: one where PBHs evaporate completely before big bang nucleosynthesis, and another where PBHs persist until today. In the case of evaporation, we analyse the emission of dark matter (DM) and dark radiation (DR) during the process of evaporation. Conversely, in the other case, the surviving PBHs themselves can contribute as DM. We further investigate how relativistic and non-relativistic accretion, together with memory burdened evaporation, impact the parameter space of the emitted DM, the abundance of stable PBHs as DM, and the contribution of DR to the effective number of relativistic degrees of freedom, $\Delta N_{\mathrm{eff}}$. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:33:54 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Maity",
"Suvashis",
""
]
] |
2507.02822 | Jiankai Sun | Wencheng Zhang, Shiqin Qiao, Lingjie Luo, Yinfeng Li, Chuanyang Zheng, Qian Xu, Meng Li, Yong Gui, Yijun He, Jianing Qiu, Jindong Hong, Jiankai Sun | SynapseRoute: An Auto-Route Switching Framework on Dual-State Large Language Model | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | With the widespread adoption of large language models (LLMs) in practical applications, selecting an appropriate model requires balancing not only performance but also operational cost. The emergence of reasoning-capable models has further widened the cost gap between "thinking" (high reasoning) and "non-thinking" (fast, low-cost) modes. In this work, we reveal that approximately 58% of medical questions can be accurately answered by the non-thinking mode alone, without requiring the high-cost reasoning process. This highlights a clear dichotomy in problem complexity and suggests that dynamically routing queries to the appropriate mode based on complexity could optimize accuracy, cost-efficiency, and overall user experience. Based on this, we further propose SynapseRoute, a machine learning-based dynamic routing framework that intelligently assigns input queries to either thinking or non-thinking modes. Experimental results on several medical datasets demonstrate that SynapseRoute not only improves overall accuracy (0.8390 vs. 0.8272) compared to the thinking mode alone but also reduces inference time by 36.8% and token consumption by 39.66%. Importantly, qualitative analysis indicates that over-reasoning on simpler queries can lead to unnecessary delays and even decreased accuracy, a pitfall avoided by our adaptive routing. Finally, this work further introduces the Accuracy-Inference-Token (AIT) index to comprehensively evaluate the trade-offs among accuracy, latency, and token cost. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:33:58 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Wencheng",
""
],
[
"Qiao",
"Shiqin",
""
],
[
"Luo",
"Lingjie",
""
],
[
"Li",
"Yinfeng",
""
],
[
"Zheng",
"Chuanyang",
""
],
[
"Xu",
"Qian",
""
],
[
"Li",
"Meng",
""
],
[
"Gui",
"Yong",
""
],
[
"He",
"Yijun",
""
],
[
"Qiu",
"Jianing",
""
],
[
"Hong",
"Jindong",
""
],
[
"Sun",
"Jiankai",
""
]
] |
2507.02823 | Luca Sodomaco | Sandra Di Rocco, Kemal Rose, Luca Sodomaco | Osculating Geometry and Higher-Order Distance Loci | 39 pages, 7 figures. Comments are welcome! | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We discuss the problem of optimizing the distance function from a given point, subject to polynomial constraints. A key algebraic invariant that governs its complexity is the Euclidean distance degree, which pertains to first-order tangency. We focus on the data locus of points possessing at least one critical point of the distance function that is normal to a higher-order osculating space. We propose a novel definition of higher-order distance degree as an intersection-theoretic invariant involving jet bundles and higher-order polar classes. Our research yields closed formulas for generic maps, Veronese embeddings, and toric embeddings. We place particular emphasis on the Bombieri-Weyl metric, revealing that the chosen metric profoundly influences both the degree and birationality of the higher-order projection maps. Additionally, we introduce a tropical framework that represents these degrees as stable intersections with Bergman fans, facilitating effective combinatorial computation in toric settings. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:34:23 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Di Rocco",
"Sandra",
""
],
[
"Rose",
"Kemal",
""
],
[
"Sodomaco",
"Luca",
""
]
] |
2507.02824 | Po-Heng Chou | Po-Heng Chou, Ching-Wen Chen, Wan-Jen Huang, Walid Saad, Yu Tsao, and Ronald Y. Chang | DNN-Based Precoding in RIS-Aided mmWave MIMO Systems With Practical Phase Shift | 5 pages, 4 figures, 2 tables, accepted by IEEE Globecom 2024 Workshops | null | null | null | eess.SP cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, the precoding design is investigated for maximizing the throughput of millimeter wave (mmWave) multiple-input multiple-output (MIMO) systems with obstructed direct communication paths. In particular, a reconfigurable intelligent surface (RIS) is employed to enhance MIMO transmissions, considering mmWave characteristics related to line-of-sight (LoS) and multipath effects. The traditional exhaustive search (ES) for optimal codewords in the continuous phase shift is computationally intensive and time-consuming. To reduce computational complexity, permuted discrete Fourier transform (DFT) vectors are used for finding codebook design, incorporating amplitude responses for practical or ideal RIS systems. However, even if the discrete phase shift is adopted in the ES, it results in significant computation and is time-consuming. Instead, the trained deep neural network (DNN) is developed to facilitate faster codeword selection. Simulation results show that the DNN maintains sub-optimal spectral efficiency even as the distance between the end-user and the RIS has variations in the testing phase. These results highlight the potential of DNN in advancing RIS-aided systems. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:35:06 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Chou",
"Po-Heng",
""
],
[
"Chen",
"Ching-Wen",
""
],
[
"Huang",
"Wan-Jen",
""
],
[
"Saad",
"Walid",
""
],
[
"Tsao",
"Yu",
""
],
[
"Chang",
"Ronald Y.",
""
]
] |
2507.02825 | Yuxuan Zhu | Yuxuan Zhu, Tengjun Jin, Yada Pruksachatkun, Andy Zhang, Shu Liu, Sasha Cui, Sayash Kapoor, Shayne Longpre, Kevin Meng, Rebecca Weiss, Fazl Barez, Rahul Gupta, Jwala Dhamala, Jacob Merizian, Mario Giulianelli, Harry Coppock, Cozmin Ududec, Jasjeet Sekhon, Jacob Steinhardt, Antony Kellerman, Sarah Schwettmann, Matei Zaharia, Ion Stoica, Percy Liang, Daniel Kang | Establishing Best Practices for Building Rigorous Agentic Benchmarks | 39 pages, 15 tables, 6 figures | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Benchmarks are essential for quantitatively tracking progress in AI. As AI agents become increasingly capable, researchers and practitioners have introduced agentic benchmarks to evaluate agents on complex, real-world tasks. These benchmarks typically measure agent capabilities by evaluating task outcomes via specific reward designs. However, we show that many agentic benchmarks have issues task setup or reward design. For example, SWE-bench Verified uses insufficient test cases, while TAU-bench counts empty responses as successful. Such issues can lead to under- or overestimation agents' performance by up to 100% in relative terms. To make agentic evaluation rigorous, we introduce the Agentic Benchmark Checklist (ABC), a set of guidelines that we synthesized from our benchmark-building experience, a survey of best practices, and previously reported issues. When applied to CVE-Bench, a benchmark with a particularly complex evaluation design, ABC reduces the performance overestimation by 33%. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:35:31 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhu",
"Yuxuan",
""
],
[
"Jin",
"Tengjun",
""
],
[
"Pruksachatkun",
"Yada",
""
],
[
"Zhang",
"Andy",
""
],
[
"Liu",
"Shu",
""
],
[
"Cui",
"Sasha",
""
],
[
"Kapoor",
"Sayash",
""
],
[
"Longpre",
"Shayne",
""
],
[
"Meng",
"Kevin",
""
],
[
"Weiss",
"Rebecca",
""
],
[
"Barez",
"Fazl",
""
],
[
"Gupta",
"Rahul",
""
],
[
"Dhamala",
"Jwala",
""
],
[
"Merizian",
"Jacob",
""
],
[
"Giulianelli",
"Mario",
""
],
[
"Coppock",
"Harry",
""
],
[
"Ududec",
"Cozmin",
""
],
[
"Sekhon",
"Jasjeet",
""
],
[
"Steinhardt",
"Jacob",
""
],
[
"Kellerman",
"Antony",
""
],
[
"Schwettmann",
"Sarah",
""
],
[
"Zaharia",
"Matei",
""
],
[
"Stoica",
"Ion",
""
],
[
"Liang",
"Percy",
""
],
[
"Kang",
"Daniel",
""
]
] |
2507.02826 | Hang Xiao | Panpan Ji, Junni Song, Hang Xiao, Hanyu Liu, Chao Li | Confidence-driven Gradient Modulation for Multimodal Human Activity Recognition: A Dynamic Contrastive Dual-Path Learning Approach | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sensor-based Human Activity Recognition (HAR) is a core technology that enables intelligent systems to perceive and interact with their environment. However, multimodal HAR systems still encounter key challenges, such as difficulties in cross-modal feature alignment and imbalanced modality contributions. To address these issues, we propose a novel framework called the Dynamic Contrastive Dual-Path Network (DCDP-HAR). The framework comprises three key components. First, a dual-path feature extraction architecture is employed, where ResNet and DenseNet branches collaboratively process multimodal sensor data. Second, a multi-stage contrastive learning mechanism is introduced to achieve progressive alignment from local perception to semantic abstraction. Third, we present a confidence-driven gradient modulation strategy that dynamically monitors and adjusts the learning intensity of each modality branch during backpropagation, effectively alleviating modality competition. In addition, a momentum-based gradient accumulation strategy is adopted to enhance training stability. We conduct ablation studies to validate the effectiveness of each component and perform extensive comparative experiments on four public benchmark datasets. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:37:46 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ji",
"Panpan",
""
],
[
"Song",
"Junni",
""
],
[
"Xiao",
"Hang",
""
],
[
"Liu",
"Hanyu",
""
],
[
"Li",
"Chao",
""
]
] |
2507.02827 | Hang Xiao | Ying Yu, Hang Xiao, Siyao Li, Jiarui Li, Haotian Tang, Hanyu Liu, Chao Li | USAD: An Unsupervised Data Augmentation Spatio-Temporal Attention Diffusion Network | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The primary objective of human activity recognition (HAR) is to infer ongoing human actions from sensor data, a task that finds broad applications in health monitoring, safety protection, and sports analysis. Despite proliferating research, HAR still faces key challenges, including the scarcity of labeled samples for rare activities, insufficient extraction of high-level features, and suboptimal model performance on lightweight devices. To address these issues, this paper proposes a comprehensive optimization approach centered on multi-attention interaction mechanisms. First, an unsupervised, statistics-guided diffusion model is employed to perform data augmentation, thereby alleviating the problems of labeled data scarcity and severe class imbalance. Second, a multi-branch spatio-temporal interaction network is designed, which captures multi-scale features of sequential data through parallel residual branches with 3*3, 5*5, and 7*7 convolutional kernels. Simultaneously, temporal attention mechanisms are incorporated to identify critical time points, while spatial attention enhances inter-sensor interactions. A cross-branch feature fusion unit is further introduced to improve the overall feature representation capability. Finally, an adaptive multi-loss function fusion strategy is integrated, allowing for dynamic adjustment of loss weights and overall model optimization. Experimental results on three public datasets, WISDM, PAMAP2, and OPPORTUNITY, demonstrate that the proposed unsupervised data augmentation spatio-temporal attention diffusion network (USAD) achieves accuracies of 98.84%, 93.81%, and 80.92% respectively, significantly outperforming existing approaches. Furthermore, practical deployment on embedded devices verifies the efficiency and feasibility of the proposed method. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:38:44 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yu",
"Ying",
""
],
[
"Xiao",
"Hang",
""
],
[
"Li",
"Siyao",
""
],
[
"Li",
"Jiarui",
""
],
[
"Tang",
"Haotian",
""
],
[
"Liu",
"Hanyu",
""
],
[
"Li",
"Chao",
""
]
] |
2507.02828 | Yuzhen Zhang | Yuzhen Zhang, Sagar Vijay, Yingfei Gu, Yimu Bao | Designs from magic-augmented Clifford circuits | 59 pages | null | null | null | quant-ph cond-mat.stat-mech cond-mat.str-el cs.IT hep-th math.IT | http://creativecommons.org/licenses/by/4.0/ | We introduce magic-augmented Clifford circuits -- architectures in which Clifford circuits are preceded and/or followed by constant-depth circuits of non-Clifford (``magic") gates -- as a resource-efficient way to realize approximate $k$-designs, with reduced circuit depth and usage of magic. We prove that shallow Clifford circuits, when augmented with constant-depth circuits of magic gates, can generate approximate unitary and state $k$-designs with $\epsilon$ relative error. The total circuit depth for these constructions on $N$ qubits is $O(\log (N/\epsilon)) +2^{O(k\log k)}$ in one dimension and $O(\log\log(N/\epsilon))+2^{O(k\log k)}$ in all-to-all circuits using ancillas, which improves upon previous results for small $k \geq 4$. Furthermore, our construction of relative-error state $k$-designs only involves states with strictly local magic. The required number of magic gates is parametrically reduced when considering $k$-designs with bounded additive error. As an example, we show that shallow Clifford circuits followed by $O(k^2)$ single-qubit magic gates, independent of system size, can generate an additive-error state $k$-design. We develop a classical statistical mechanics description of our random circuit architectures, which provides a quantitative understanding of the required depth and number of magic gates for additive-error state $k$-designs. We also prove no-go theorems for various architectures to generate designs with bounded relative error. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:41:03 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Yuzhen",
""
],
[
"Vijay",
"Sagar",
""
],
[
"Gu",
"Yingfei",
""
],
[
"Bao",
"Yimu",
""
]
] |
2507.02829 | Allen Zang | Allen Zang, Tian-Xing Zheng, Peter C. Maurer, Frederic T. Chong, Martin Suchara, and Tian Zhong | Enhancing Noisy Quantum Sensing by GHZ State Partitioning | 32 pages, 13 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Presence of harmful noise is inevitable in entanglement-enhanced sensing systems, requiring careful allocation of resources to optimize sensing performance in practical scenarios. We advocate a simple but effective strategy to improve sensing performance in the presence of noise. Given a fixed number of quantum sensors, we partition the preparation of GHZ states by preparing smaller, independent sub-ensembles of GHZ states instead of a GHZ state across all sensors. We perform extensive analytical studies of the phase estimation performance when using partitioned GHZ states under realistic noise -- including state preparation error, particle loss during parameter encoding, and sensor dephasing during parameter encoding. We derive simple, closed-form expressions that quantify the optimal number of sub-ensembles for partitioned GHZ states. We also examine the explicit noisy quantum sensing dynamics under dephasing and loss, where we demonstrate the advantage from partitioning for maximal QFI, short-time QFI increase, and the sensing performance in the sequential scheme. The results offer quantitative insights into the sensing performance impact of different noise sources and reinforce the importance of resource allocation optimization in realistic quantum applications. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:42:23 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zang",
"Allen",
""
],
[
"Zheng",
"Tian-Xing",
""
],
[
"Maurer",
"Peter C.",
""
],
[
"Chong",
"Frederic T.",
""
],
[
"Suchara",
"Martin",
""
],
[
"Zhong",
"Tian",
""
]
] |
2507.02830 | Yitong Yao | Yitong Yao, Jun Li, Jiacheng Ye, Fan Yang, Dao-Xin Yao | Electric Field Induced Superconductivity in Bilayer Octagraphene | 7 pages, 6 figures | Chin. Phys. Lett., 2025, 42(6): 067504 | 10.1088/0256-307X/42/6/067504 | null | cond-mat.supr-con | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the energy bands, magnetism, and superconductivity of bilayer octagraphene with A-A stacking under a perpendicular electric field. A tight-binding model is used to analyze the band structure of the system. The doubling of the unit cell results in each band of the single layer splitting into two. We find that applying a perpendicular electric field increases the band splitting. As the electric field strength increases, the nesting of the Fermi Surface(FS) weakens, eventually disrupting the antiferromagnetic order and bilayer octagraphene exhibits superconductivity. Spin fluctuations can induce unconventional superconductivity with s+--wave pairing. Applying a perpendicular electric field to bilayer octagraphene parent weakens the nesting of the FS, ultimately killing the spin-density-wave (SDW) ordered state and transitioning it into the superconducting state, whichworks as a doping effect. We use the random-phase approximation approach to obtain the pairing eigenvalues and pairing symmetries of the perpendicular electric field-tuned bilayer octagraphene in the weak coupling limit. By tuning the strength of the perpendicular electric field, the critical interaction strength for SDW order can be modified, which in turn may promote the emergence of unconventional superconductivity. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:42:45 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yao",
"Yitong",
""
],
[
"Li",
"Jun",
""
],
[
"Ye",
"Jiacheng",
""
],
[
"Yang",
"Fan",
""
],
[
"Yao",
"Dao-Xin",
""
]
] |
2507.02831 | Fabrizio Nieri | Fabrizio Nieri | Trace Formulas for Deformed W-Algebras | 14 pages, 1 figure | null | null | null | hep-th math.QA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate trace formulas in $\varepsilon$-deformed W-algebras, highlighting a novel connection to the modular double of $q$-deformed W-algebras. In particular, we show that torus correlators in the additive (Yangian) setting reproduce sphere correlators in the trigonometric setup, possibly with the inclusion of a non-perturbative completion. From a dual perspective, this mechanism implements a gauge theoretic 2d$\to$3d uplift, where a circle direction in the world-sheet transmutes to a compact space-time direction in a non-trivial manner. We further discuss a unified picture of deformed W-algebras driven by trace formulas, suggesting a deeper algebraic layer related to the massive and massless form-factor approach to integrable QFT and 2d CFT. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:42:54 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Nieri",
"Fabrizio",
""
]
] |
2507.02832 | Hongshun Yao | Hongshun Yao, Xia Liu, Mingrui Jing, and Xin Wang | LCQNN: Linear Combination of Quantum Neural Networks | null | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | Quantum neural networks combine quantum computing with advanced data-driven methods, offering promising applications in quantum machine learning. However, the optimal paradigm of quantum neural networks in the context trainability and expressivity remains an open question. To overcome this issue, we first introduce a framework called Linear Combination of Quantum Neural Networks (LCQNN), which leverages the linear combination of unitaries concept to coordinate the composition of QNNs, thereby balancing trainability and expressivity. Secondly, LCQNN provides a tunable design that can reduce vanishing gradients without incurring excessive classical simulability. Specifically, we detail how restricting certain subspaces or adopting $k$-local control unitaries prevents gradients from collapsing while maintaining enough parameter volume for complex tasks. These findings also align with prior discussions on expanding QNN capabilities for multi-qubit systems. Additionally, we extend the LCQNN to the group action scenarios. By exploiting symmetry, the LCQNN model, excluding exponentially large irreducible subspaces, can circumvent barren plateaus. Overall, LCQNN provides a novel framework for focusing quantum resources into architectures that remain practically trainable yet expressive enough to tackle challenging machine-learning applications. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:43:10 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yao",
"Hongshun",
""
],
[
"Liu",
"Xia",
""
],
[
"Jing",
"Mingrui",
""
],
[
"Wang",
"Xin",
""
]
] |
2507.02833 | Valentina Pyatkin | Valentina Pyatkin and Saumya Malik and Victoria Graf and Hamish Ivison and Shengyi Huang and Pradeep Dasigi and Nathan Lambert and Hannaneh Hajishirzi | Generalizing Verifiable Instruction Following | 11 pages | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | A crucial factor for successful human and AI interaction is the ability of language models or chatbots to follow human instructions precisely. A common feature of instructions are output constraints like ``only answer with yes or no" or ``mention the word `abrakadabra' at least 3 times" that the user adds to craft a more useful answer. Even today's strongest models struggle with fulfilling such constraints. We find that most models strongly overfit on a small set of verifiable constraints from the benchmarks that test these abilities, a skill called precise instruction following, and are not able to generalize well to unseen output constraints. We introduce a new benchmark, IFBench, to evaluate precise instruction following generalization on 58 new, diverse, and challenging verifiable out-of-domain constraints. In addition, we perform an extensive analysis of how and on what data models can be trained to improve precise instruction following generalization. Specifically, we carefully design constraint verification modules and show that reinforcement learning with verifiable rewards (RLVR) significantly improves instruction following. In addition to IFBench, we release 29 additional new hand-annotated training constraints and verification functions, RLVR training prompts, and code. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:44:33 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Pyatkin",
"Valentina",
""
],
[
"Malik",
"Saumya",
""
],
[
"Graf",
"Victoria",
""
],
[
"Ivison",
"Hamish",
""
],
[
"Huang",
"Shengyi",
""
],
[
"Dasigi",
"Pradeep",
""
],
[
"Lambert",
"Nathan",
""
],
[
"Hajishirzi",
"Hannaneh",
""
]
] |
2507.02834 | Ruiyang Zhou | Ruiyang Zhou, Shuozhe Li, Amy Zhang, Liu Leqi | ExPO: Unlocking Hard Reasoning with Self-Explanation-Guided Reinforcement Learning | null | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in large language models have been driven by reinforcement learning (RL)-style post-training, which improves reasoning by optimizing model outputs based on reward or preference signals. GRPO-style approaches implement this by using self-generated samples labeled by an outcome-based verifier. However, these methods depend heavily on the model's initial ability to produce positive samples. They primarily refine what the model already knows (distribution sharpening) rather than enabling the model to solve problems where it initially fails. This limitation is especially problematic in early-stage RL training and on challenging reasoning tasks, where positive samples are unlikely to be generated. To unlock reasoning ability in such settings, the model must explore new reasoning trajectories beyond its current output distribution. Such exploration requires access to sufficiently good positive samples to guide the learning. While expert demonstrations seem like a natural solution, we find that they are often ineffective in RL post-training. Instead, we identify two key properties of effective positive samples: they should (1) be likely under the current policy, and (2) increase the model's likelihood of predicting the correct answer. Based on these insights, we propose $\textbf{Self-Explanation Policy Optimization (ExPO)}$-a simple and modular framework that generates such samples by conditioning on the ground-truth answer. ExPO enables efficient exploration and guides the model to produce reasoning trajectories more aligned with its policy than expert-written CoTs, while ensuring higher quality than its own (incorrect) samples. Experiments show that ExPO improves both learning efficiency and final performance on reasoning benchmarks, surpassing expert-demonstration-based methods in challenging settings such as MATH level-5, where the model initially struggles the most. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:44:55 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhou",
"Ruiyang",
""
],
[
"Li",
"Shuozhe",
""
],
[
"Zhang",
"Amy",
""
],
[
"Leqi",
"Liu",
""
]
] |
2507.02835 | Biswajit Pandey | Amit Mondal, Biswajit Pandey, Krishna Ghosh | Revealing a transitional epoch of large-scale cosmic anisotropy in the quasar distribution | 17 pages, 1 table, 10 figures, Codes are available on GitHub at https://github.com/amit280695/renyi-entropy-quasars | null | null | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Cosmological Principle posits that the Universe is isotropic on the largest scales. While widely supported, this foundational assumption remains testable. We analyse the angular distribution of over one million quasars from the Gaia-unWISE catalogue using Renyi entropy, a multiscale statistical measure sensitive to higher-order clustering. Dividing the sample into three redshift bins, we find that both the low- and high-redshift distributions are statistically consistent with isotropy. However, at intermediate redshift ($1 \leq z < 2.2$), we detect a statistically significant and scale-dependent anisotropy that persists under stringent masking, suggesting a physical origin. We interpret this as evidence for a transitional epoch in cosmic history, during which large-scale structures such as superclusters became prominent before their growth was gradually damped by the onset of accelerated expansion. These findings position Renyi entropy as a powerful probe of cosmic evolution and highlight the potential thermodynamic links between structure formation, entropy dissipation, and the emergence of large-scale isotropy. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:45:18 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Mondal",
"Amit",
""
],
[
"Pandey",
"Biswajit",
""
],
[
"Ghosh",
"Krishna",
""
]
] |
2507.02836 | Daniela De Silva | Daniela De Silva and Ovidiu Savin | On the Boundary Harnack Principle for operators with different lower order terms | null | null | null | null | math.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We provide the classical Boundary Harnack principle in Lipschitz domains for solutions to two different linear uniformly elliptic equations with the same principal part. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:45:27 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"De Silva",
"Daniela",
""
],
[
"Savin",
"Ovidiu",
""
]
] |
2507.02837 | Matteo Carducci | Giulia Bevilacqua, Matteo Carducci | Free boundary regularity for a tumor growth model with obstacle | null | null | null | null | math.AP math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop an existence and regularity theory for solutions to a geometric free boundary problem motivated by models of tumor growth. In this setting, the tumor invades an accessible region $D$, its motion is directed along a constant vector $V$, and it cannot penetrate another region $K$ acting as an obstacle to the spread of the tumor. Due to the non variational structure of the problem, we show existence of viscosity solutions via Perron's method. Subsequently, we prove interior regularity for the free boundary near regular points by means of an improvement of flatness argument. We further analyze the boundary regularity and we prove that the free boundary meets the obstacle as a $C^{1,\alpha}$ graph. A key step in the analysis of the boundary regularity involves the study of a thin obstacle problem with oblique boundary conditions, for which we establish $C^{1,\alpha}$ estimates. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:45:46 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Bevilacqua",
"Giulia",
""
],
[
"Carducci",
"Matteo",
""
]
] |
2507.02838 | Michal P. Heller | Michal P. Heller and Clemens Werthmann | Early time hydrodynamic attractor in a nearly-unitary Fermi gas | 8 pages, 3 figures | null | null | null | hep-th cond-mat.quant-gas hep-ph nucl-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The hydrodynamic attractors paradigm aims to explain fast applicability of hydrodynamics at ultrarelativistic nuclear collisions at RHIC and LHC in terms of the emergence of a universal behaviour across different initial states. The attractors observed in theoretical models are first driven by a rapid expansion of matter, which later gives way to the decay of exponentially damped transient excitations into the viscous fluid regime. Recently a theoretical proposal was made how to realize hydrodynamic attractors in cold atomic gases focusing on the decay of transients and opening prospects for a future experimental realization in table top experiments. Motivated by this, in the present paper we show how the far-from-equilibrium, expansion-driven part of the hydrodynamic attractor emerges in a model of a nearly-unitary Fermi gas dynamics under full theoretical control. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:46:03 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Heller",
"Michal P.",
""
],
[
"Werthmann",
"Clemens",
""
]
] |
2507.02839 | Lek-Heng Lim | Zehua Lai, Lek-Heng Lim, and Tianyun Tang | Stiefel optimization is NP-hard | 9 pages | null | null | null | math.OC cs.CC | http://creativecommons.org/licenses/by/4.0/ | We show that linearly constrained linear optimization over a Stiefel or Grassmann manifold is NP-hard in general. We show that the same is true for unconstrained quadratic optimization over a Stiefel manifold. We will establish the nonexistence of FPTAS for these optimization problems over a Stiefel manifold. As an aside we extend our results to flag manifolds. Combined with earlier findings, this shows that manifold optimization is a difficult endeavor -- even the simplest problems like LP and unconstrained QP are already NP-hard on the most common manifolds. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:47:22 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Lai",
"Zehua",
""
],
[
"Lim",
"Lek-Heng",
""
],
[
"Tang",
"Tianyun",
""
]
] |
2507.02840 | CaiChang Li | Li-Na Yan, Xiang-Yan Gao, Gao-Da Liu, Cai-Chang Li | Neutrino mixing parameters and masses from $\Delta(96)\rtimes H_{CP}$ in the tri-direct CP approach | 48 pages, 10 figures | null | null | null | hep-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a comprehensive model independent analysis of all breaking patterns resulting from $\Delta(96)\rtimes H_{CP}$ in the tri-direct CP approach of the minimal seesaw model with two right-handed neutrinos. The three generations of left-handed lepton doublets are assumed to transform as the irreducible triplet $\bm{3_{0}}$ of $\Delta(96)$, and the two right-handed neutrinos are assigned to singlets. In the case that both flavon fields $\phi_{\text{atm}}$ and $\phi_{\text{sol}}$ transform as triplet $\bm{\bar{3}_{0}}$, only one phenomenologically viable lepton mixing pattern is obtained for normal ordering neutrino masses. The lepton mixing matrix is predicted to be TM1 pattern, with neutrino masses, mixing angles, and CP violation phases depending on only three real input parameters. When $\phi_{\text{sol}}$ is assigned to the $\bm{\bar{3}_{1}}$ representation, an additional real parameter $x$ must be included. Then we find 42 independent phenomenologically interesting mixing patterns, and the corresponding predictions for lepton mixing parameters and neutrino masses are obtained. Furthermore, we perform a detailed numerical analysis for five example breaking patterns with some benchmark values of $x$. The absolute values of the first columns of PMNS matrix are fixed to be $\left(\sqrt{\frac{2}{3}},\frac{1}{\sqrt{6}},\frac{1}{\sqrt{6}}\right)^{T}$, $\frac{1}{5}\left(\sqrt{17},2,2\right)^{T}$, $\frac{1}{\sqrt{38}}\left(5,2,3\right)^{T}$, $\frac{1}{\sqrt{57}}\left(\sqrt{37},\sqrt{10},\sqrt{10}\right)^{T}$ and $\frac{1}{3}\left(\sqrt{6},1,\sqrt{2}\right)^{T}$, respectively. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:48:21 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Yan",
"Li-Na",
""
],
[
"Gao",
"Xiang-Yan",
""
],
[
"Liu",
"Gao-Da",
""
],
[
"Li",
"Cai-Chang",
""
]
] |
2507.02841 | Kaiyi Zhang | Kaiyi Zhang, Ang Lv, Jinpeng Li, Yongbo Wang, Feng Wang, Haoyuan Hu, Rui Yan | StepHint: Multi-level Stepwise Hints Enhance Reinforcement Learning to Reason | null | null | null | null | cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Reinforcement learning with verifiable rewards (RLVR) is a promising approach for improving the complex reasoning abilities of large language models (LLMs). However, current RLVR methods face two significant challenges: the near-miss reward problem, where a small mistake can invalidate an otherwise correct reasoning process, greatly hindering training efficiency; and exploration stagnation, where models tend to focus on solutions within their ``comfort zone,'' lacking the motivation to explore potentially more effective alternatives. To address these challenges, we propose StepHint, a novel RLVR algorithm that utilizes multi-level stepwise hints to help models explore the solution space more effectively. StepHint generates valid reasoning chains from stronger models and partitions these chains into reasoning steps using our proposed adaptive partitioning method. The initial few steps are used as hints, and simultaneously, multiple-level hints (each comprising a different number of steps) are provided to the model. This approach directs the model's exploration toward a promising solution subspace while preserving its flexibility for independent exploration. By providing hints, StepHint mitigates the near-miss reward problem, thereby improving training efficiency. Additionally, the external reasoning pathways help the model develop better reasoning abilities, enabling it to move beyond its ``comfort zone'' and mitigate exploration stagnation. StepHint outperforms competitive RLVR enhancement methods across six mathematical benchmarks, while also demonstrating superior generalization and excelling over baselines on out-of-domain benchmarks. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:51:06 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Kaiyi",
""
],
[
"Lv",
"Ang",
""
],
[
"Li",
"Jinpeng",
""
],
[
"Wang",
"Yongbo",
""
],
[
"Wang",
"Feng",
""
],
[
"Hu",
"Haoyuan",
""
],
[
"Yan",
"Rui",
""
]
] |
2507.02842 | Justin Chen | Anders Aamand, Maryam Aliakbarpour, Justin Y. Chen, Shyam Narayanan, Sandeep Silwal | On the Structure of Replicable Hypothesis Testers | Abstract abridged to meet arxiv requirements | null | null | null | cs.DS | http://creativecommons.org/licenses/by/4.0/ | A hypothesis testing algorithm is replicable if, when run on two different samples from the same distribution, it produces the same output with high probability. This notion, defined by by Impagliazzo, Lei, Pitassi, and Sorell [STOC'22], can increase trust in testing procedures and is deeply related to algorithmic stability, generalization, and privacy. We build general tools to prove lower and upper bounds on the sample complexity of replicable testers, unifying and quantitatively improving upon existing results.
We identify a set of canonical properties, and prove that any replicable testing algorithm can be modified to satisfy these properties without worsening accuracy or sample complexity. A canonical replicable algorithm computes a deterministic function of its input (i.e., a test statistic) and thresholds against a uniformly random value in $[0,1]$. It is invariant to the order in which the samples are received, and, if the testing problem is ``symmetric,'' then the algorithm is also invariant to the labeling of the domain elements, resolving an open question by Liu and Ye [NeurIPS'24]. We prove new lower bounds for uniformity, identity, and closeness testing by reducing to the case where the replicable algorithm satisfies these canonical properties.
We systematize and improve upon a common strategy for replicable algorithm design based on test statistics with known expectation and bounded variance. Our framework allow testers which have been extensively analyzed in the non-replicable setting to be made replicable with minimal overhead. As direct applications of our framework, we obtain constant-factor optimal bounds for coin testing and closeness testing and get replicability for free in a large parameter regime for uniformity testing.
We also give state-of-the-art bounds for replicable Gaussian mean testing, and, unlike prior work, our algorithm runs in polynomial time. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:51:31 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Aamand",
"Anders",
""
],
[
"Aliakbarpour",
"Maryam",
""
],
[
"Chen",
"Justin Y.",
""
],
[
"Narayanan",
"Shyam",
""
],
[
"Silwal",
"Sandeep",
""
]
] |
2507.02843 | Yuchen Ma | Yuchen Ma, Dennis Frauen, Jonas Schweisthal, Stefan Feuerriegel | LLM-Driven Treatment Effect Estimation Under Inference Time Text Confounding | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Estimating treatment effects is crucial for personalized decision-making in medicine, but this task faces unique challenges in clinical practice. At training time, models for estimating treatment effects are typically trained on well-structured medical datasets that contain detailed patient information. However, at inference time, predictions are often made using textual descriptions (e.g., descriptions with self-reported symptoms), which are incomplete representations of the original patient information. In this work, we make three contributions. (1) We show that the discrepancy between the data available during training time and inference time can lead to biased estimates of treatment effects. We formalize this issue as an inference time text confounding problem, where confounders are fully observed during training time but only partially available through text at inference time. (2) To address this problem, we propose a novel framework for estimating treatment effects that explicitly accounts for inference time text confounding. Our framework leverages large language models together with a custom doubly robust learner to mitigate biases caused by the inference time text confounding. (3) Through a series of experiments, we demonstrate the effectiveness of our framework in real-world applications. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:52:27 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Ma",
"Yuchen",
""
],
[
"Frauen",
"Dennis",
""
],
[
"Schweisthal",
"Jonas",
""
],
[
"Feuerriegel",
"Stefan",
""
]
] |
2507.02844 | Yi Ding | Ziqi Miao, Yi Ding, Lijun Li, Jing Shao | Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection | 16 pages | null | null | null | cs.CV cs.CL cs.CR | http://creativecommons.org/licenses/by/4.0/ | With the emergence of strong visual-language capabilities, multimodal large language models (MLLMs) have demonstrated tremendous potential for real-world applications. However, the security vulnerabilities exhibited by the visual modality pose significant challenges to deploying such models in open-world environments. Recent studies have successfully induced harmful responses from target MLLMs by encoding harmful textual semantics directly into visual inputs. However, in these approaches, the visual modality primarily serves as a trigger for unsafe behavior, often exhibiting semantic ambiguity and lacking grounding in realistic scenarios. In this work, we define a novel setting: visual-centric jailbreak, where visual information serves as a necessary component in constructing a complete and realistic jailbreak context. Building on this setting, we propose the VisCo (Visual Contextual) Attack. VisCo fabricates contextual dialogue using four distinct visual-focused strategies, dynamically generating auxiliary images when necessary to construct a visual-centric jailbreak scenario. To maximize attack effectiveness, it incorporates automatic toxicity obfuscation and semantic refinement to produce a final attack prompt that reliably triggers harmful responses from the target black-box MLLMs. Specifically, VisCo achieves a toxicity score of 4.78 and an Attack Success Rate (ASR) of 85% on MM-SafetyBench against GPT-4o, significantly outperforming the baseline, which performs a toxicity score of 2.48 and an ASR of 22.2%. The code is available at https://github.com/Dtc7w3PQ/Visco-Attack. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:53:12 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Miao",
"Ziqi",
""
],
[
"Ding",
"Yi",
""
],
[
"Li",
"Lijun",
""
],
[
"Shao",
"Jing",
""
]
] |
2507.02845 | Davide Giordano Ario Altamura | Davide Giordano Ario Altamura, Jos\'e Luis Gaona-Reyes, Elliot Simcox, Hendrik Ulbricht, Angelo Bassi | Enhancement of the effects due to the Schr\"odinger-Newton equation | null | null | null | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Schr\"odinger-Newton (SN) equation introduces a nonlinear self-gravitational term to the standard Schr\"odinger equation, offering a paradigmatic model for semiclassical gravity. However, the small deviations it predicts from standard quantum mechanics pose significant experimental challenges. We propose a novel method to amplify such deviations through periodic modulation of the trapping frequency in a levitated mechanical oscillator. We identify specific regimes where the SN-induced effects on the dynamics of second moments are significantly enhanced-by up to six orders of magnitude compared to unmodulated setups. We show that this protocol remains feasible within current magnetic levitation technologies and enables distinguishability between standard and SN dynamics using measurable quantities such as the position variance. Our results pave the way for a viable experimental test of the SN equation, offering a new route to probe the interface between quantum mechanics and gravity. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:53:28 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Altamura",
"Davide Giordano Ario",
""
],
[
"Gaona-Reyes",
"José Luis",
""
],
[
"Simcox",
"Elliot",
""
],
[
"Ulbricht",
"Hendrik",
""
],
[
"Bassi",
"Angelo",
""
]
] |
2507.02846 | Anmol Singhal | Anmol Singhal, Travis Breaux | Legal Requirements Translation from Law | 13 pages, 7 figures, Accepted at the 33rd IEEE International Requirements Engineering 2025 | null | null | null | cs.SE cs.CL | http://creativecommons.org/licenses/by/4.0/ | Software systems must comply with legal regulations, which is a resource-intensive task, particularly for small organizations and startups lacking dedicated legal expertise. Extracting metadata from regulations to elicit legal requirements for software is a critical step to ensure compliance. However, it is a cumbersome task due to the length and complex nature of legal text. Although prior work has pursued automated methods for extracting structural and semantic metadata from legal text, key limitations remain: they do not consider the interplay and interrelationships among attributes associated with these metadata types, and they rely on manual labeling or heuristic-driven machine learning, which does not generalize well to new documents. In this paper, we introduce an approach based on textual entailment and in-context learning for automatically generating a canonical representation of legal text, encodable and executable as Python code. Our representation is instantiated from a manually designed Python class structure that serves as a domain-specific metamodel, capturing both structural and semantic legal metadata and their interrelationships. This design choice reduces the need for large, manually labeled datasets and enhances applicability to unseen legislation. We evaluate our approach on 13 U.S. state data breach notification laws, demonstrating that our generated representations pass approximately 89.4% of test cases and achieve a precision and recall of 82.2 and 88.7, respectively. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:53:48 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Singhal",
"Anmol",
""
],
[
"Breaux",
"Travis",
""
]
] |
2507.02847 | Kunyu Zhang | Kunyu Zhang, Qiang Li, and Shujian Yu | MvHo-IB: Multi-View Higher-Order Information Bottleneck for Brain Disorder Diagnosis | Accepted by MICCAI-25, code is available at \url{https://github.com/zky04/MvHo-IB} | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent evidence suggests that modeling higher-order interactions (HOIs) in functional magnetic resonance imaging (fMRI) data can enhance the diagnostic accuracy of machine learning systems. However, effectively extracting and utilizing HOIs remains a significant challenge. In this work, we propose MvHo-IB, a novel multi-view learning framework that integrates both pairwise interactions and HOIs for diagnostic decision-making, while automatically compressing task-irrelevant redundant information. MvHo-IB introduces several key innovations: (1) a principled method that combines O-information from information theory with a matrix-based Renyi alpha-order entropy estimator to quantify and extract HOIs, (2) a purpose-built Brain3DCNN encoder to effectively utilize these interactions, and (3) a new multi-view learning information bottleneck objective to enhance representation learning. Experiments on three benchmark fMRI datasets demonstrate that MvHo-IB achieves state-of-the-art performance, significantly outperforming previous methods, including recent hypergraph-based techniques. The implementation of MvHo-IB is available at https://github.com/zky04/MvHo-IB. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:54:03 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhang",
"Kunyu",
""
],
[
"Li",
"Qiang",
""
],
[
"Yu",
"Shujian",
""
]
] |
2507.02848 | Xiao Han | Xiao Han and Shahn Majid | Quantum jet Hopf algebroids by cotwist | 28 pages Latex | null | null | null | math.QA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a cotwist construction for Hopf algebroids that also entails cotwisting or `quantisation' of the base and which is dual to a previous twisting construction of Xu. Whereas the latter applied the construction to the algebra of differential operators on a classical base $B$, we show that the dual of this is the algebra of sections $J^\infty(B)$ of the jet bundle and hence the latter forms a Hopf algebroid, which we identify as a quotient of the pair Hopf algebroid $B\otimes B$. This classical jet bundle is then quantised by our cotwist construction to give a noncommutative jet Hopf algebroid over a noncommutative base. We also observe in the commutative case that $J^k(B)$ for jets of order $k$ can be identified with $J^1(B_k)$ where $B_k$ denotes $B$ equipped with a certain non-standard first order differential calculus. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:54:55 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Han",
"Xiao",
""
],
[
"Majid",
"Shahn",
""
]
] |
2507.02849 | Usha Devi A R Dr. | H. Talath, B. P. Govindaraja, B. G. Divyamani, Akshata Shenoy H., A. R. Usha Devi, and Sudha | Three-qubit W state tomography via full and marginal state reconstructions on ibm_osaka | 8 pages, 5 figures | null | null | null | quant-ph | http://creativecommons.org/licenses/by/4.0/ | We present a three-qubit quantum state tomography scheme requiring a set of 17 measurements, significantly reducing the experimental overhead compared to the conventional 63 Pauli measurements. Using IBM's 127-qubit open-access quantum processor ibm osaka, we prepare the three-qubit W state and employ our tomography scheme to reconstruct it. Additionally, we implement a two-qubit tomography protocol, involving a set of 7 measurements, on ibm osaka to reconstruct two of the two-qubit marginals of the W state. This serves as a proof-of-principle demonstration of the well-known theoretical result that any two of the two-qubit reduced density matrices can uniquely determine most of the whole three-qubit pure states. We show that the fidelity of the W-state reconstructed from its two-qubit subsystems is consistently larger than that obtained from the full three-qubit tomography, highlighting the practical advantage of the subsystem-based tomography approach. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:55:19 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Talath",
"H.",
""
],
[
"Govindaraja",
"B. P.",
""
],
[
"Divyamani",
"B. G.",
""
],
[
"H.",
"Akshata Shenoy",
""
],
[
"Devi",
"A. R. Usha",
""
],
[
"Sudha",
"",
""
]
] |
2507.02850 | Leshem Choshen | Almog Hilel, Idan Shenfeld, Leshem Choshen, Jacob Andreas | LLM Hypnosis: Exploiting User Feedback for Unauthorized Knowledge Injection to All Users | null | null | null | null | cs.CL cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a vulnerability in language models (LMs) trained with user feedback, whereby a single user can persistently alter LM knowledge and behavior given only the ability to provide prompts and upvote / downvote feedback on LM outputs. To implement the attack, the attacker prompts the LM to stochastically output either a "poisoned" or benign response, then upvotes the poisoned response or downvotes the benign one. When feedback signals are used in a subsequent preference tuning behavior, LMs exhibit increased probability of producing poisoned responses even in contexts without malicious prompts. We show that this attack can be used to (1) insert factual knowledge the model did not previously possess, (2) modify code generation patterns in ways that introduce exploitable security flaws, and (3) inject fake financial news. Our finding both identifies a new qualitative feature of language model preference tuning (showing that it even highly restricted forms of preference data can be used to exert fine-grained control over behavior), and a new attack mechanism for LMs trained with user feedback (extending work on pretraining-time data poisoning and deployment-time prompt injection). | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:55:40 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Hilel",
"Almog",
""
],
[
"Shenfeld",
"Idan",
""
],
[
"Choshen",
"Leshem",
""
],
[
"Andreas",
"Jacob",
""
]
] |
2507.02851 | Purbesh Mitra | Purbesh Mitra and Sennur Ulukus | MOTIF: Modular Thinking via Reinforcement Fine-tuning in LLMs | null | null | null | null | cs.CL cs.AI cs.IT cs.LG cs.SY eess.SY math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in the reasoning capabilities of large language models (LLMs) show that employing group relative policy optimization (GRPO) algorithm for reinforcement learning (RL) training allows the models to use more thinking/reasoning tokens for generating better responses. However, LLMs can generate only a finite amount of tokens while maintaining attention to the previously generated tokens. This limit, also known as the context size of an LLM, is a bottleneck in LLM reasoning with arbitrarily large number of tokens. To think beyond the limit of context size, an LLM must employ a modular thinking strategy to reason over multiple rounds. In this work, we propose $\textbf{MOTIF: Modular Thinking via Reinforcement Finetuning}$ -- an RL training method for generating thinking tokens in multiple rounds, effectively allowing the model to think with additional context size. We trained the open-source model Qwen2.5-3B-Instruct on GSM8K dataset via parameter efficient fine-tuning and tested its accuracy on MATH500 and AIME2024 benchmarks. Our experiments show 3.8\% and 3.3\% improvements over vanilla GRPO based training in the respective benchmarks. Furthermore, this improvement was achieved with only 15\% of samples, thus demonstrating sample efficiency of MOTIF. Our code and models are available at https://github.com/purbeshmitra/MOTIF and https://huggingface.co/purbeshmitra/MOTIF, respectively. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:55:43 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Mitra",
"Purbesh",
""
],
[
"Ulukus",
"Sennur",
""
]
] |
2507.02852 | Martijn Kool | M. Kool and J.V. Rennemo | Proof of a magnificent conjecture | 59 pages | null | null | null | math.AG hep-th math.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motivated by super-Yang-Mills theory on a Calabi-Yau 4-fold, Nekrasov and Piazzalunga have assigned weights to $r$-tuples of solid partitions and conjectured a formula for their weighted generating function.
We define $K$-theoretic virtual invariants of Quot schemes of 0-dimensional quotients of $\mathcal{O}_{\mathbb{C}^4}^{\oplus r}$ by realizing them as zero loci of isotropic sections of orthogonal bundles on non-commutative Quot schemes. Via the Oh-Thomas localization formula, we recover Nekrasov-Piazzalunga's weights and derive their sign rule.
Our proof passes through refining the $K$-theoretic invariants to sheaves and describing them via Clifford modules, which lets us show that they arise from a factorizable sequence of sheaves in the sense of Okounkov. Taking limits of the equivariant parameters, we then deduce the Nekrasov-Piazzalunga conjecture from its 3-dimensional analog. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:55:52 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Kool",
"M.",
""
],
[
"Rennemo",
"J. V.",
""
]
] |
2507.02853 | Bikram Pain | Bikram Pain, Ratul Thakur, Sthitadhi Roy | Imprints of information scrambling on eigenstates of a quantum chaotic system | 8 pages, 3 figures + supplementary material (6 pages, 2 figures) | null | null | null | quant-ph cond-mat.dis-nn cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | How are the spatial and temporal patterns of information scrambling in locally interacting quantum many-body systems imprinted on the eigenstates of the system's time-evolution operator? We address this question by identifying statistical correlations among sets of minimally four eigenstates that provide a unified framework for various measures of information scrambling. These include operator mutual information and operator entanglement entropy of the time-evolution operator, as well as more conventional diagnostics such as two-point dynamical correlations and out-of-time-ordered correlators. We demonstrate this framework by deriving exact results for eigenstate correlations in a minimal model of quantum chaos -- Floquet dual-unitary circuits. These results reveal not only the butterfly effect and the information lightcone, but also finer structures of scrambling within the lightcone. Our work thus shows how the eigenstates of a chaotic system can encode the full spatiotemporal anatomy of quantum chaos, going beyond the descriptions offered by random matrix theory and the eigenstate thermalisation hypothesis. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:57:23 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Pain",
"Bikram",
""
],
[
"Thakur",
"Ratul",
""
],
[
"Roy",
"Sthitadhi",
""
]
] |
2507.02854 | Daniel Campbell PhD | Daniel Campbell, Luigi D'Onofrio, Tom\'a\v{s} V\'itek | Diffeomorphic approximation of piecewise affine homeomorphisms | null | null | null | null | math.AP | http://creativecommons.org/licenses/by/4.0/ | Given any $f$ a locally finitely piecewise affine homeomorphism of $\Omega \subset \mathbb{R}^d$ onto $\Delta \subset \mathbb{R}^d$ (for $d=3, 4$) such that $f\in W^{1,p}(\Omega, \mathbb{R}^d)$ and $f^{-1}\in W^{1,q}(\Delta, \mathbb{R}^d)$, $1\leq p ,q < \infty$ and any $\epsilon >0$ we construct a diffeomorphism $\tilde{f}$ such that
$$\|f-\tilde{f}\|_{W^{1,p}(\Omega,\mathbb{R}^d)} + \|f^{-1}-\tilde{f}^{-1}\|_{W^{1,q}(\Delta,\mathbb{R}^d)} < \epsilon.$$ | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:58:38 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Campbell",
"Daniel",
""
],
[
"D'Onofrio",
"Luigi",
""
],
[
"Vítek",
"Tomáš",
""
]
] |
2507.02855 | Colin Rothgang | Colin Rothgang and Florian Rabe | Subtyping in DHOL -- Extended preprint | 16 pages main document, 44 pages of appendices, to be published in FroCoS 2025 | null | null | null | cs.LO cs.AI cs.FL | http://creativecommons.org/licenses/by/4.0/ | The recently introduced dependent typed higher-order logic (DHOL) offers an interesting compromise between expressiveness and automation support. It sacrifices the decidability of its type system in order to significantly extend its expressiveness over standard HOL. Yet it retains strong automated theorem proving support via a sound and complete translation to HOL.
We leverage this design to extend DHOL with refinement and quotient types. Both of these are commonly requested by practitioners but rarely provided by automated theorem provers. This is because they inherently require undecidable typing and thus are very difficult to retrofit to decidable type systems. But with DHOL already doing the heavy lifting, adding them is not only possible but elegant and simple.
Concretely, we add refinement and quotient types as special cases of subtyping. This turns the associated canonical inclusion resp. projection maps into identity maps and thus avoids costly changes in representation. We present the syntax, semantics, and translation to HOL for the extended language, including the proofs of soundness and completeness. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:00 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Rothgang",
"Colin",
""
],
[
"Rabe",
"Florian",
""
]
] |
2507.02856 | Shashwat Goel | Nikhil Chandak and Shashwat Goel and Ameya Prabhu and Moritz Hardt and Jonas Geiping | Answer Matching Outperforms Multiple Choice for Language Model Evaluation | 34 pages, Code is available at https://github.com/nikhilchandak/answer-matching | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multiple choice benchmarks have long been the workhorse of language model evaluation because grading multiple choice is objective and easy to automate. However, we show multiple choice questions from popular benchmarks can often be answered without even seeing the question. These shortcuts arise from a fundamental limitation of discriminative evaluation not shared by evaluations of the model's free-form, generative answers. Until recently, there appeared to be no viable, scalable alternative to multiple choice--but, we show that this has changed. We consider generative evaluation via what we call answer matching: Give the candidate model the question without the options, have it generate a free-form response, then use a modern language model with the reference answer to determine if the response matches the reference. To compare the validity of different evaluation strategies, we annotate MMLU-Pro and GPQA-Diamond to obtain human grading data, and measure the agreement of each evaluation approach. We find answer matching using recent models--even small ones--achieves near-perfect agreement, in the range of inter-annotator agreement. In contrast, both multiple choice evaluation and using LLM-as-a-judge without reference answers aligns poorly with human grading. Improving evaluations via answer matching is not merely a conceptual concern: the rankings of several models change significantly when evaluating their free-form responses with answer matching. In light of these findings, we discuss how to move the evaluation ecosystem from multiple choice to answer matching. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:02 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Chandak",
"Nikhil",
""
],
[
"Goel",
"Shashwat",
""
],
[
"Prabhu",
"Ameya",
""
],
[
"Hardt",
"Moritz",
""
],
[
"Geiping",
"Jonas",
""
]
] |
2507.02857 | Henghui Ding | Ziye Li, Hao Luo, Xincheng Shuai, Henghui Ding | AnyI2V: Animating Any Conditional Image with Motion Control | ICCV 2025, Project Page: https://henghuiding.com/AnyI2V/ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in video generation, particularly in diffusion models, have driven notable progress in text-to-video (T2V) and image-to-video (I2V) synthesis. However, challenges remain in effectively integrating dynamic motion signals and flexible spatial constraints. Existing T2V methods typically rely on text prompts, which inherently lack precise control over the spatial layout of generated content. In contrast, I2V methods are limited by their dependence on real images, which restricts the editability of the synthesized content. Although some methods incorporate ControlNet to introduce image-based conditioning, they often lack explicit motion control and require computationally expensive training. To address these limitations, we propose AnyI2V, a training-free framework that animates any conditional images with user-defined motion trajectories. AnyI2V supports a broader range of modalities as the conditional image, including data types such as meshes and point clouds that are not supported by ControlNet, enabling more flexible and versatile video generation. Additionally, it supports mixed conditional inputs and enables style transfer and editing via LoRA and text prompts. Extensive experiments demonstrate that the proposed AnyI2V achieves superior performance and provides a new perspective in spatial- and motion-controlled video generation. Code is available at https://henghuiding.com/AnyI2V/. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:02 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Li",
"Ziye",
""
],
[
"Luo",
"Hao",
""
],
[
"Shuai",
"Xincheng",
""
],
[
"Ding",
"Henghui",
""
]
] |
2507.02858 | Anmol Singhal | Yuchen Shen, Anmol Singhal, Travis Breaux | Requirements Elicitation Follow-Up Question Generation | 13 pages, 2 figures, accepted at the 33rd IEEE International Requirements Engineering 2025 | null | null | null | cs.SE cs.CL | http://creativecommons.org/licenses/by/4.0/ | Interviews are a widely used technique in eliciting requirements to gather stakeholder needs, preferences, and expectations for a software system. Effective interviewing requires skilled interviewers to formulate appropriate interview questions in real time while facing multiple challenges, including lack of familiarity with the domain, excessive cognitive load, and information overload that hinders how humans process stakeholders' speech. Recently, large language models (LLMs) have exhibited state-of-the-art performance in multiple natural language processing tasks, including text summarization and entailment. To support interviewers, we investigate the application of GPT-4o to generate follow-up interview questions during requirements elicitation by building on a framework of common interviewer mistake types. In addition, we describe methods to generate questions based on interviewee speech. We report a controlled experiment to evaluate LLM-generated and human-authored questions with minimal guidance, and a second controlled experiment to evaluate the LLM-generated questions when generation is guided by interviewer mistake types. Our findings demonstrate that, for both experiments, the LLM-generated questions are no worse than the human-authored questions with respect to clarity, relevancy, and informativeness. In addition, LLM-generated questions outperform human-authored questions when guided by common mistakes types. This highlights the potential of using LLMs to help interviewers improve the quality and ease of requirements elicitation interviews in real time. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:04 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shen",
"Yuchen",
""
],
[
"Singhal",
"Anmol",
""
],
[
"Breaux",
"Travis",
""
]
] |
2507.02859 | Jiaer Xia | Jiaer Xia, Bingkui Tong, Yuhang Zang, Rui Shao, Kaiyang Zhou | Bootstrapping Grounded Chain-of-Thought in Multimodal LLMs for Data-Efficient Model Adaptation | Accepted by ICCV2025 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities in interpreting images using natural language. However, without using large-scale datasets for retraining, these models are difficult to adapt to specialized vision tasks, e.g., chart understanding. This problem is caused by a mismatch between pre-training and downstream datasets: pre-training datasets primarily concentrate on scenes and objects but contain limited information about specialized, non-object images, such as charts and tables. In this paper, we share an interesting finding that training an MLLM with Chain-of-Thought (CoT) reasoning data can facilitate model adaptation in specialized vision tasks, especially under data-limited regimes. However, we identify a critical issue within CoT data distilled from pre-trained MLLMs, i.e., the data often contains multiple factual errors in the reasoning steps. To address the problem, we propose Grounded Chain-of-Thought (GCoT), a simple bootstrapping-based approach that aims to inject grounding information (i.e., bounding boxes) into CoT data, essentially making the reasoning steps more faithful to input images. We evaluate our approach on five specialized vision tasks, which cover a variety of visual formats including charts, tables, receipts, and reports. The results demonstrate that under data-limited regimes our approach significantly improves upon fine-tuning and distillation. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:29 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Xia",
"Jiaer",
""
],
[
"Tong",
"Bingkui",
""
],
[
"Zang",
"Yuhang",
""
],
[
"Shao",
"Rui",
""
],
[
"Zhou",
"Kaiyang",
""
]
] |
2507.02860 | Xin Zhou | Xin Zhou, Dingkang Liang, Kaijin Chen, Tianrui Feng, Xiwu Chen, Hongkai Lin, Yikang Ding, Feiyang Tan, Hengshuang Zhao, Xiang Bai | Less is Enough: Training-Free Video Diffusion Acceleration via Runtime-Adaptive Caching | The code is made available at https://github.com/H-EmbodVis/EasyCache. Project page: https://h-embodvis.github.io/EasyCache/ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video generation models have demonstrated remarkable performance, yet their broader adoption remains constrained by slow inference speeds and substantial computational costs, primarily due to the iterative nature of the denoising process. Addressing this bottleneck is essential for democratizing advanced video synthesis technologies and enabling their integration into real-world applications. This work proposes EasyCache, a training-free acceleration framework for video diffusion models. EasyCache introduces a lightweight, runtime-adaptive caching mechanism that dynamically reuses previously computed transformation vectors, avoiding redundant computations during inference. Unlike prior approaches, EasyCache requires no offline profiling, pre-computation, or extensive parameter tuning. We conduct comprehensive studies on various large-scale video generation models, including OpenSora, Wan2.1, and HunyuanVideo. Our method achieves leading acceleration performance, reducing inference time by up to 2.1-3.3$\times$ compared to the original baselines while maintaining high visual fidelity with a significant up to 36% PSNR improvement compared to the previous SOTA method. This improvement makes our EasyCache a efficient and highly accessible solution for high-quality video generation in both research and practical applications. The code is available at https://github.com/H-EmbodVis/EasyCache. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:54 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Zhou",
"Xin",
""
],
[
"Liang",
"Dingkang",
""
],
[
"Chen",
"Kaijin",
""
],
[
"Feng",
"Tianrui",
""
],
[
"Chen",
"Xiwu",
""
],
[
"Lin",
"Hongkai",
""
],
[
"Ding",
"Yikang",
""
],
[
"Tan",
"Feiyang",
""
],
[
"Zhao",
"Hengshuang",
""
],
[
"Bai",
"Xiang",
""
]
] |
2507.02861 | Zhening Huang | Zhening Huang, Xiaoyang Wu, Fangcheng Zhong, Hengshuang Zhao, Matthias Nie{\ss}ner, Joan Lasenby | LiteReality: Graphics-Ready 3D Scene Reconstruction from RGB-D Scans | Project Page: https://litereality.github.io; Video: https://www.youtube.com/watch?v=ecK9m3LXg2c&feature=youtu.be | null | null | null | cs.CV cs.AI cs.GR | http://creativecommons.org/licenses/by/4.0/ | We propose LiteReality, a novel pipeline that converts RGB-D scans of indoor environments into compact, realistic, and interactive 3D virtual replicas. LiteReality not only reconstructs scenes that visually resemble reality but also supports key features essential for graphics pipelines -- such as object individuality, articulation, high-quality physically based rendering materials, and physically based interaction. At its core, LiteReality first performs scene understanding and parses the results into a coherent 3D layout and objects with the help of a structured scene graph. It then reconstructs the scene by retrieving the most visually similar 3D artist-crafted models from a curated asset database. Next, the Material Painting module enhances realism by recovering high-quality, spatially varying materials. Finally, the reconstructed scene is integrated into a simulation engine with basic physical properties to enable interactive behavior. The resulting scenes are compact, editable, and fully compatible with standard graphics pipelines, making them suitable for applications in AR/VR, gaming, robotics, and digital twins. In addition, LiteReality introduces a training-free object retrieval module that achieves state-of-the-art similarity performance on the Scan2CAD benchmark, along with a robust material painting module capable of transferring appearances from images of any style to 3D assets -- even under severe misalignment, occlusion, and poor lighting. We demonstrate the effectiveness of LiteReality on both real-life scans and public datasets. Project page: https://litereality.github.io; Video: https://www.youtube.com/watch?v=ecK9m3LXg2c | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:55 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Huang",
"Zhening",
""
],
[
"Wu",
"Xiaoyang",
""
],
[
"Zhong",
"Fangcheng",
""
],
[
"Zhao",
"Hengshuang",
""
],
[
"Nießner",
"Matthias",
""
],
[
"Lasenby",
"Joan",
""
]
] |
2507.02862 | Xiang Fan | Xiang Fan, Xiaohang Sun, Kushan Thakkar, Zhu Liu, Vimal Bhat, Ranjay Krishna, Xiang Hao | RefTok: Reference-Based Tokenization for Video Generation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Effectively handling temporal redundancy remains a key challenge in learning video models. Prevailing approaches often treat each set of frames independently, failing to effectively capture the temporal dependencies and redundancies inherent in videos. To address this limitation, we introduce RefTok, a novel reference-based tokenization method capable of capturing complex temporal dynamics and contextual information. Our method encodes and decodes sets of frames conditioned on an unquantized reference frame. When decoded, RefTok preserves the continuity of motion and the appearance of objects across frames. For example, RefTok retains facial details despite head motion, reconstructs text correctly, preserves small patterns, and maintains the legibility of handwriting from the context. Across 4 video datasets (K600, UCF-101, BAIR Robot Pushing, and DAVIS), RefTok significantly outperforms current state-of-the-art tokenizers (Cosmos and MAGVIT) and improves all evaluated metrics (PSNR, SSIM, LPIPS) by an average of 36.7% at the same or higher compression ratios. When a video generation model is trained using RefTok's latents on the BAIR Robot Pushing task, the generations not only outperform MAGVIT-B but the larger MAGVIT-L, which has 4x more parameters, across all generation metrics by an average of 27.9%. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:55 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Fan",
"Xiang",
""
],
[
"Sun",
"Xiaohang",
""
],
[
"Thakkar",
"Kushan",
""
],
[
"Liu",
"Zhu",
""
],
[
"Bhat",
"Vimal",
""
],
[
"Krishna",
"Ranjay",
""
],
[
"Hao",
"Xiang",
""
]
] |
2507.02863 | Wenzhao Zheng | Yuqi Wu, Wenzhao Zheng, Jie Zhou, Jiwen Lu | Point3R: Streaming 3D Reconstruction with Explicit Spatial Pointer Memory | Code is available at: https://github.com/YkiWu/Point3R | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dense 3D scene reconstruction from an ordered sequence or unordered image collections is a critical step when bringing research in computer vision into practical scenarios. Following the paradigm introduced by DUSt3R, which unifies an image pair densely into a shared coordinate system, subsequent methods maintain an implicit memory to achieve dense 3D reconstruction from more images. However, such implicit memory is limited in capacity and may suffer from information loss of earlier frames. We propose Point3R, an online framework targeting dense streaming 3D reconstruction. To be specific, we maintain an explicit spatial pointer memory directly associated with the 3D structure of the current scene. Each pointer in this memory is assigned a specific 3D position and aggregates scene information nearby in the global coordinate system into a changing spatial feature. Information extracted from the latest frame interacts explicitly with this pointer memory, enabling dense integration of the current observation into the global coordinate system. We design a 3D hierarchical position embedding to promote this interaction and design a simple yet effective fusion mechanism to ensure that our pointer memory is uniform and efficient. Our method achieves competitive or state-of-the-art performance on various tasks with low training costs. Code is available at: https://github.com/YkiWu/Point3R. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:56 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Wu",
"Yuqi",
""
],
[
"Zheng",
"Wenzhao",
""
],
[
"Zhou",
"Jie",
""
],
[
"Lu",
"Jiwen",
""
]
] |
2507.02864 | Renhao Wang | Renhao Wang, Haoran Geng, Tingle Li, Feishi Wang, Gopala Anumanchipalli, Philipp Wu, Trevor Darrell, Boyi Li, Pieter Abbeel, Jitendra Malik, Alexei A. Efros | MultiGen: Using Multimodal Generation in Simulation to Learn Multimodal Policies in Real | null | null | null | null | cs.RO cs.CV | http://creativecommons.org/licenses/by/4.0/ | Robots must integrate multiple sensory modalities to act effectively in the real world. Yet, learning such multimodal policies at scale remains challenging. Simulation offers a viable solution, but while vision has benefited from high-fidelity simulators, other modalities (e.g. sound) can be notoriously difficult to simulate. As a result, sim-to-real transfer has succeeded primarily in vision-based tasks, with multimodal transfer still largely unrealized. In this work, we tackle these challenges by introducing MultiGen, a framework that integrates large-scale generative models into traditional physics simulators, enabling multisensory simulation. We showcase our framework on the dynamic task of robot pouring, which inherently relies on multimodal feedback. By synthesizing realistic audio conditioned on simulation video, our method enables training on rich audiovisual trajectories -- without any real robot data. We demonstrate effective zero-shot transfer to real-world pouring with novel containers and liquids, highlighting the potential of generative modeling to both simulate hard-to-model modalities and close the multimodal sim-to-real gap. | [
{
"version": "v1",
"created": "Thu, 03 Jul 2025 17:59:58 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Wang",
"Renhao",
""
],
[
"Geng",
"Haoran",
""
],
[
"Li",
"Tingle",
""
],
[
"Wang",
"Feishi",
""
],
[
"Anumanchipalli",
"Gopala",
""
],
[
"Wu",
"Philipp",
""
],
[
"Darrell",
"Trevor",
""
],
[
"Li",
"Boyi",
""
],
[
"Abbeel",
"Pieter",
""
],
[
"Malik",
"Jitendra",
""
],
[
"Efros",
"Alexei A.",
""
]
] |
cond-mat/9307047 | Sriram Shastry | B. Sriram Shastry (AT& T Bell Laboratories, Murray Hill, New Jersey), Elihu Abrahams (Serin Physics Laboratory, Rutgers University, Piscataway, New Jersey) | What Does The Korringa Ratio Measure? | 24 June 1993 | Phys. Rev. Letts. 72, 1933 (1994) | 10.1103/PhysRevLett.72.1933 | null | cond-mat | http://creativecommons.org/licenses/by/4.0/ | We present an analysis of the Korringa ratio in a dirty metal, emphasizing the case where a Stoner enhancement of the uniform susceptibilty is present. We find that the relaxation rates are significantly enhanced by disorder, and that the inverse problem of determining the bare density of states from a study of the change of the Knight shift and relaxation rates with some parameter, such as pressure, has rather constrained solutions, with the disorder playing an important role. Some preliminary applications to the case of chemical substitution in the Rb$_{3-x}$K$_x $C$_{60}$ family of superconductors is presented and some other relevant systems are mentioned. | [
{
"version": "v1",
"created": "Wed, 21 Jul 1993 17:36:11 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 23:23:30 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shastry",
"B. Sriram",
"",
"AT& T Bell Laboratories, Murray Hill, New Jersey"
],
[
"Abrahams",
"Elihu",
"",
"Serin Physics Laboratory, Rutgers University, Piscataway, New Jersey"
]
] |
cond-mat/9402065 | Sriram Shastry | B. Sriram Shastry (AT\&T Bell Labs, Murray Hill, NJ) | The $1/r^2$ Integrable system: The Universal Hamiltonian for Quantum Chaos | [email protected], Published in the proceedings of the 16th Taniguchi International Symposium on the Theory of Condensed Matter: "Correlation Effects in Low Dimensional Electron Systems'', Springer Verlag 1994 eds. N. Kawakami and A. Okiji | null | null | null | cond-mat hep-th | http://creativecommons.org/licenses/by/4.0/ | We summarize recent work showing that the $1/r^2$ model of interacting particles in 1-dimension is a universal Hamiltonian for quantum chaotic systems. The problem is analyzed in terms of random matrices and of the evolution of their eigenvalues under changes of parameters. The robustness of bulk space-time correlations of a many particle system to changing boundary conditions is suggested to be at the root of the universality. The explicit density-density correlation functions of the $1/r^2$ model, now available through the above mapping at two values of the coupling constant, are interpreted in the light of Bethe's {\it Ansatz}, giving a vivid picture of the fractionalization of bare particles or holes into ``quark'' like Bethe quasi-particles and holes. | [
{
"version": "v1",
"created": "Wed, 16 Feb 1994 22:23:50 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 23:35:07 GMT"
}
] | 2025-07-04T00:00:00 | [
[
"Shastry",
"B. Sriram",
"",
"AT\\&T Bell Labs, Murray Hill, NJ"
]
] |
1110.2912 | Tanusri Saha-Dasgupta | Hena Das, G. Sangiovanni, A. Valli, K. Held and T. Saha-Dasgupta | Size control of Charge-Orbital Order in Half-Doped Manganite, La$_{0.5}$Ca$_{0.5}$MnO$_3$ | 5 pages, 3 figures; to appear in Physical Review Letters | Phys. Rev. Lett. 107, 197202 (2011) | 10.1103/PhysRevLett.107.197202 | null | cond-mat.str-el cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motivated by recent experimental results, we study the effect of size reduction on half-doped manganite, La$_{0.5}$Ca$_{0.5}$MnO$_3$, using the combination of density functional theory (DFT) and dynamical mean field theory (DMFT). We find that upon size reduction, the charge-ordered antiferromagnetic phase, observed in bulk, to be destabilized, giving rise to the stability of a ferromagnetic metallic state. Our theoretical results, carried out on defect-free nanocluster in isolation, establish the structural changes that follow upon size reduction to be responsible for this. Our study further points out the effect of size reduction to be distinctively different from application of hydrostatic pressure. Interestingly, our DFT+DMFT study, additionally, reports the correlation-driven stability of charge-orbitally ordered state in bulk La$_{0.5}$Ca$_{0.5}$MnO$_3$, even in absence of long range magnetic order. | [
{
"version": "v1",
"created": "Thu, 13 Oct 2011 12:20:23 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Das",
"Hena",
""
],
[
"Sangiovanni",
"G.",
""
],
[
"Valli",
"A.",
""
],
[
"Held",
"K.",
""
],
[
"Saha-Dasgupta",
"T.",
""
]
] |
1509.03875 | Kondo Gnanvo | Kondo Gnanvo, Xinzhan Bai, Chao Gua, Nilanga Liyanage, Vladimir Nelyubin, Yuxiang Zhao | Performance in Test Beam of a Large-area and Light-weight GEM detector with 2D Stereo-Angle (U-V) Strip Readout | null | Nuclear Inst. and Methods in Physics Research, A 58313, 2015 | 10.1016/j.nima.2015.11.071 | null | physics.ins-det | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A large-area and light-weight Gas Electron Multiplier (GEM) detector was built at the University of Virginia as a prototype for the detector R$\&$D program of the future Electron Ion Collider. The prototype has a trapezoidal geometry designed as a generic sector module in a disk layer configuration of a forward tracker in collider detectors. It is based on light-weight material and narrow support frames in order to minimize multiple scattering and dead-to-sensitive area ratio. The chamber has a novel type of two dimensional (2D) stereo-angle readout board with U-V strips that provides (r,$\varphi$) position information in the cylindrical coordinate system of a collider environment. The prototype was tested at the Fermilab Test Beam Facility in October 2013 and the analysis of the test beam data demonstrates an excellent response uniformity of the large area chamber with an efficiency higher than 95%. An angular resolution of 60 $\mu$rad in the azimuthal direction and a position resolution better than 550 $\mu$m in the radial direction were achieved with the U-V strip readout board. The results are discussed in this paper. | [
{
"version": "v1",
"created": "Sun, 13 Sep 2015 18:12:00 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Gnanvo",
"Kondo",
""
],
[
"Bai",
"Xinzhan",
""
],
[
"Gua",
"Chao",
""
],
[
"Liyanage",
"Nilanga",
""
],
[
"Nelyubin",
"Vladimir",
""
],
[
"Zhao",
"Yuxiang",
""
]
] |
1706.03716 | Wenfei Liu | Wenfei Liu | The minimal volume of log surfaces of general type with positive geometric genus | 42 pages, to appear in Israel Journal of Mathematics | null | null | null | math.AG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We determine the minimal possible volume of a projective log canonical surface of general type with prescribed positive geometric genus. As applications, we provide effecitive Noether type inequalities for log canonical threefolds and stable surfaces. Also, we obtain a uniform bound on the order of the symplectic automorphism group $\mathrm{Aut}_s(S)$ of smooth projective surfaces $S$ of general type. | [
{
"version": "v1",
"created": "Mon, 12 Jun 2017 16:27:34 GMT"
},
{
"version": "v2",
"created": "Tue, 23 Aug 2022 09:24:32 GMT"
},
{
"version": "v3",
"created": "Wed, 02 Jul 2025 02:40:56 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Liu",
"Wenfei",
""
]
] |
1706.08329 | Christoph Wernhard | Christoph Wernhard | The Boolean Solution Problem from the Perspective of Predicate Logic -- Extended Version | v4: Removed an erroneous and a sketchy section; minor corrections | null | 10.1007/978-3-319-66167-4_19 | null | cs.LO cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Finding solution values for unknowns in Boolean equations was a principal reasoning mode in the Algebra of Logic of the 19th century. Schr\"oder investigated it as Aufl\"osungsproblem (solution problem). It is closely related to the modern notion of Boolean unification. Today it is commonly presented in an algebraic setting, but seems potentially useful also in knowledge representation based on predicate logic. We show that it can be modeled on the basis of first-order logic extended by second-order quantification. A wealth of classical results transfers, foundations for algorithms unfold, and connections with second-order quantifier elimination and Craig interpolation become apparent. Although for first-order inputs the set of solutions is recursively enumerable, the development of constructive methods remains a challenge. We identify some cases that allow constructions, most of them based on Craig interpolation. | [
{
"version": "v1",
"created": "Mon, 26 Jun 2017 11:30:06 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Oct 2017 09:19:48 GMT"
},
{
"version": "v3",
"created": "Mon, 18 Dec 2017 22:29:44 GMT"
},
{
"version": "v4",
"created": "Wed, 02 Jul 2025 10:31:26 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Wernhard",
"Christoph",
""
]
] |
1907.10604 | Keiji Matsumoto | Keiji Matsumoto | Reversibility of distance measures of states with some focus on total variation distance | null | null | null | null | quant-ph math-ph math.MP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Consider a classical system, which is in the state described by probability distribution $p$ or $q$, and embed these classical informations into quantum system by a physical map $\Gamma$, $\rho=\Gamma(p)$ and $\sigma=\Gamma(q)$. Intuitively, the pair $\{p_{\rho}^{M},p_{\sigma}^{M}\}$ of the distributions of the data of the measurement $M$ on the pair $\{\rho,\sigma\}$ should contain strictly less information than the pair $\{p,q\}$ provided the pair $\{\rho,\sigma\}$ is non-commutative. Indeed, this statement had been shown if the information is measured by $f$-divergence such that $f$ is operator convex. In the paper, the statement is extended to the case where $f$ is strictly convex. Also, we disprove the assertion for the total variation distance $\Vert p-q\Vert_{1}$, the $f$-divergence with $f(r)=|1-r|$: if $\{\rho,\sigma\}$ satisfies some not very restrictive conditions, $\Vert p_{\rho}^{M}-p_{\sigma}^{M}\Vert_{1}$ equals $\Vert p-q\Vert_{1}$. Here we present sufficient condition for general case, and necessary and sufficient condition for qubit states. | [
{
"version": "v1",
"created": "Wed, 24 Jul 2019 15:50:11 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Matsumoto",
"Keiji",
""
]
] |
1912.00459 | Samuel Bismuth | Samuel Bismuth, Ivan Bliznets, Erel Segal-Halevi | Fair Division with Bounded Sharing: Binary and Non-Degenerate Valuations | Full version of a paper accepted to SAGT 2024 | null | null | null | cs.GT econ.TH | http://creativecommons.org/licenses/by/4.0/ | A set of objects is to be divided fairly among agents with different tastes, modeled by additive utility-functions. If we consider the objects as indivisible, many instances of the decision problem: ``Is there a fair division of the objects among the agents'' are negative. In addition, this question is hard to solve even for most of the special cases. The latter reasons give us a good motivation to relax the problem for which the running time complexity is better, and the number of positive instances (admitting a fair division) will significantly grow. Whereas many works relax the fairness criteria, this paper introduces another relaxation: an agent is allowed to share a \emph{bounded} number of objects between two or more agents in order to attain fairness. The paper studies various notions of fairness, such as proportionality, envy-freeness, equitability, and consensus. We analyze the run-time complexity of finding a fair allocation with a given number of sharings under several restrictions on the agents' valuations, such as: binary, generalized-binary, and non-degenerate. | [
{
"version": "v1",
"created": "Sun, 01 Dec 2019 18:11:13 GMT"
},
{
"version": "v2",
"created": "Tue, 16 Jul 2024 13:22:35 GMT"
},
{
"version": "v3",
"created": "Wed, 02 Jul 2025 13:26:30 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Bismuth",
"Samuel",
""
],
[
"Bliznets",
"Ivan",
""
],
[
"Segal-Halevi",
"Erel",
""
]
] |
2002.02232 | Sam Slezak | Elizabeth Crosson, Samuel Slezak | Classical Simulation of High Temperature Quantum Ising Models | 10 pages, 1 figure | null | null | null | quant-ph cond-mat.stat-mech cs.DS | http://creativecommons.org/licenses/by/4.0/ | We consider generalized quantum Ising models, including those which could describe disordered materials or quantum annealers, and we prove that for all temperatures above a system-size independent threshold the path integral Monte Carlo method based on worldline heat-bath updates always mixes to stationarity in time $\mathcal{O}(n \log n)$ for an $n$ qubit system, and therefore provides a fully polynomial-time approximation scheme for the partition function. This result holds whenever the temperature is greater than four plus twice the maximum interaction degree (valence) over all qubits, measured in units of the local coupling strength. For example, this implies that the classical simulation of the thermal state of a superconducting device modeling a frustrated quantum Ising model with maximum valence of 6 and coupling strengths of 1 GHz is always possible at temperatures above 800 mK. Despite the quantum system being at high temperature, the classical spin system resulting from the quantum-to-classical mapping contains strong couplings which cause the single-site Glauber dynamics to mix slowly, therefore this result depends on the use of worldline updates (which are a form of cluster updates that can be implemented efficiently). This result places definite constraints on the temperatures required for a quantum advantage in analog quantum simulation with various NISQ devices based on equilibrium states of quantum Ising models. | [
{
"version": "v1",
"created": "Thu, 06 Feb 2020 12:56:42 GMT"
},
{
"version": "v2",
"created": "Tue, 01 Jul 2025 22:36:17 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Crosson",
"Elizabeth",
""
],
[
"Slezak",
"Samuel",
""
]
] |
2003.05355 | Igor Mikol\'a\v{s}ek | Igor Mikolasek | New stochastic highway capacity estimation method and why product limit method is unsuitable | Withdrawn due to invalid methodology conclusions and replaced with arXiv:2507.00893 | null | null | null | stat.AP stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Kaplan-Meier estimate, commonly known as product limit method (PLM), and maximum likelihood estimate (MLE) methods in general are often cited as means of stochastic highway capacity estimation. This article discusses their unsuitability for such application as properties of traffic flow do not meet the assumptions for use of the methods. They assume the observed subject has a history which it went through and did not fail. However, due to its nature, each traffic flow measurement behaves as a separate subject which did not go through all the lower levels of intensity (did not "age"). An alternative method is proposed. It fits the resulting cumulative frequency of breakdowns with respect to the traffic flow intensity leading to the breakdown instead of directly estimating the underlying probability distribution of capacity. Analyses of accuracy and sensitivity to data quantity and censoring rate of the new method are provided along with comparison to the PLM. The results prove unsuitability of the PLM and MLE methods in general. The new method is then used in a case study which compares capacity of a work-zone with and without a traffic flow speed harmonisation system installed. The results confirm positive effect of harmonisation on capacity. | [
{
"version": "v1",
"created": "Wed, 11 Mar 2020 15:25:58 GMT"
},
{
"version": "v2",
"created": "Wed, 02 Jul 2025 06:11:28 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mikolasek",
"Igor",
""
]
] |
2004.08643 | Alessandro Portaluri | Xijun Hu, Alessandro Portaluri, Li Wu, Qin Xing | Morse index theorem for heteroclinic, homoclinic and halfclinic orbits of Lagrangian systems | 38 pages, 1 figure | null | null | null | math.DS math.CA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The purpose of this paper is to prove a new, more general version of the Morse index theorem for heteroclinic, homoclinic, and half-clinic solutions in general Lagrangian systems. In the final section, we compute the Morse index for specific heteroclinic and half-clinic solutions in classical mechanical models such as the mathematical pendulum, the Nagumo equation, and a four-dimensional competition-diffusion system. | [
{
"version": "v1",
"created": "Sat, 18 Apr 2020 15:30:09 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Jun 2024 18:53:11 GMT"
},
{
"version": "v3",
"created": "Wed, 02 Jul 2025 10:07:27 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Hu",
"Xijun",
""
],
[
"Portaluri",
"Alessandro",
""
],
[
"Wu",
"Li",
""
],
[
"Xing",
"Qin",
""
]
] |
2006.13052 | Alexander Patkowski | Alexander E. Patkowski | On asymptotic expansions for basic hypergeometric functions | Added details to the proof of Theorem 1.1 | null | null | null | math.NT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper establishes new results concerning asymptotic expansions of $q$-series related to partial theta functions. We first establish a new method to obtain asymptotic expansions using a result of Ono and Lovejoy, and then build on these observations to obtain asymptotic expansions for related multi hypergeometric series. | [
{
"version": "v1",
"created": "Tue, 23 Jun 2020 14:25:04 GMT"
},
{
"version": "v2",
"created": "Tue, 01 Apr 2025 02:01:29 GMT"
},
{
"version": "v3",
"created": "Wed, 02 Jul 2025 01:12:25 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Patkowski",
"Alexander E.",
""
]
] |
2006.16824 | Primoz Skraba | Primoz Skraba, Katharine Turner | Wasserstein Stability for Persistence Diagrams | This is now only the cellular stability and applications. For the algebraic part see previous version as it will be made into a standalone paper shortly | null | null | null | math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The stability of persistence diagrams is among the most important results in applied and computational topology. Most results in the literature phrase stability in terms of the bottleneck distance between diagrams and the $\infty$-norm of perturbations. This has two main implications: it makes the space of persistence diagrams rather pathological and it is often provides very pessimistic bounds with respect to outliers. In this paper, we provide new stability results with respect to the $p$-Wasserstein distance between persistence diagrams. This includes an elementary proof for the setting of functions on sufficiently finite spaces in terms of the $p$-norm of the perturbations, along with an algebraic framework for $p$-Wasserstein distance which extends the results to wider class of modules. We also provide apply the results to a wide range of applications in topological data analysis (TDA) including topological summaries, persistence transforms and the special but important case of Vietoris-Rips complexes. | [
{
"version": "v1",
"created": "Tue, 30 Jun 2020 14:01:05 GMT"
},
{
"version": "v2",
"created": "Mon, 09 Nov 2020 10:32:26 GMT"
},
{
"version": "v3",
"created": "Thu, 04 Mar 2021 09:42:32 GMT"
},
{
"version": "v4",
"created": "Tue, 29 Nov 2022 09:28:39 GMT"
},
{
"version": "v5",
"created": "Wed, 20 Sep 2023 15:05:44 GMT"
},
{
"version": "v6",
"created": "Wed, 02 Jul 2025 09:05:39 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Skraba",
"Primoz",
""
],
[
"Turner",
"Katharine",
""
]
] |
2007.06691 | Taimur Mohammadi | Taimur Mohammadi, Behrooz Malekolkalami | The Power Spectrum Of Gravitational Waves in Anisotropic Universe (\emph{Bianchi type--I}) | 7 pages, 3 figure. arXiv admin note: substantial text overlap with arXiv:1912.12474 | null | null | null | gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One of the predictions from simple inflation models is stochastic background of gravitational waves (or literally what is called Primordial Gravitational Waves (\textbf{PGW})) with a nearly scale--invariant spectrum. To discuss a possible direct detection of PGW, the quantity so--called Spectral Energy Density (\textbf{SED}) has crucial role. In this work, we consider PGW amplified by Radiation dominate and generated by perturbing the Anisotropic (\emph{Bianchi type--I}) metric. We focus on the SED quantity of these gravity waves. The intended frameworks are the Einstein's and modified (pure quadratic \emph{Ricci} scalar) gravities. The spectral graphs for SED shows that the power spectrum of gravitational waves in the early universe in the isotropic universe was different from the isotropic universe, but at the present time (age of universe) they coincide. | [
{
"version": "v1",
"created": "Mon, 13 Jul 2020 21:06:23 GMT"
},
{
"version": "v2",
"created": "Sun, 13 Sep 2020 19:27:15 GMT"
},
{
"version": "v3",
"created": "Sun, 18 Jul 2021 14:53:18 GMT"
},
{
"version": "v4",
"created": "Wed, 17 Nov 2021 17:30:51 GMT"
},
{
"version": "v5",
"created": "Tue, 30 Nov 2021 20:08:31 GMT"
},
{
"version": "v6",
"created": "Sat, 05 Feb 2022 20:08:29 GMT"
},
{
"version": "v7",
"created": "Wed, 13 Jul 2022 20:35:02 GMT"
},
{
"version": "v8",
"created": "Wed, 02 Jul 2025 08:54:31 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Mohammadi",
"Taimur",
""
],
[
"Malekolkalami",
"Behrooz",
""
]
] |
2011.04524 | Markus Szymik | Victoria Lebed (LMNO), Markus Szymik (University of Sheffield and NTNU) | The homology of permutation racks | 19 pages, Math. Z. (to appear) | Math. Z. (2025) 311:5 | 10.1007/s00209-025-03797-5 | null | math.AT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite a blossoming of research activity on racks and their homology for over two decades, with a record of diverse applications to central parts of contemporary mathematics, there are still very few examples of racks whose homology has been fully calculated. In this paper, we compute the entire integral homology of all permutation racks. Our method of choice involves homotopical algebra, which was brought to bear on the homology of racks only recently. For our main result, we establish a spectral sequence, which reduces the problem to one in equivariant homology, and for which we show that it always degenerates. The blueprint given in this paper demonstrates the high potential for further exploitation of these techniques. | [
{
"version": "v1",
"created": "Mon, 09 Nov 2020 16:03:43 GMT"
},
{
"version": "v2",
"created": "Tue, 27 May 2025 16:35:26 GMT"
}
] | 2025-07-03T00:00:00 | [
[
"Lebed",
"Victoria",
"",
"LMNO"
],
[
"Szymik",
"Markus",
"",
"University of Sheffield and NTNU"
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.