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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2506.04146 | Simone Cantori | Simone Cantori, Andrea Mari, David Vitali, Sebastiano Pilati | Deep-learned error mitigation via partially knitted circuits for the variational quantum eigensolver | 19 pages, 9 figures | null | null | null | quant-ph cond-mat.dis-nn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The variational quantum eigensolver (VQE) is generally regarded as a promising quantum algorithm for near-term noisy quantum computers. However, when implemented with the deep circuits that are in principle required for achieving a satisfactory accuracy, the algorithm is strongly limited by noise. Here, we show how to make VQE functional via a tailored error mitigation technique based on deep learning. Our method employs multilayer perceptrons trained on the fly to predict ideal expectation values from noisy outputs combined with circuit descriptors. Importantly, a circuit knitting technique with partial knitting is adopted to substantially reduce the classical computational cost of creating the training data. We also show that other popular general-purpose quantum error mitigation techniques do not reach comparable accuracies. Our findings highlight the power of deep-learned quantum error mitigation methods tailored to specific circuit families, and of the combined use of variational quantum algorithms and classical deep learning. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:40:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cantori",
"Simone",
""
],
[
"Mari",
"Andrea",
""
],
[
"Vitali",
"David",
""
],
[
"Pilati",
"Sebastiano",
""
]
] |
2506.04147 | Jiaheng Hu | Jiaheng Hu, Peter Stone, Roberto Mart\'in-Mart\'in | SLAC: Simulation-Pretrained Latent Action Space for Whole-Body Real-World RL | null | null | null | null | cs.RO cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Building capable household and industrial robots requires mastering the control of versatile, high-degree-of-freedom (DoF) systems such as mobile manipulators. While reinforcement learning (RL) holds promise for autonomously acquiring robot control policies, scaling it to high-DoF embodiments remains challenging. Direct RL in the real world demands both safe exploration and high sample efficiency, which are difficult to achieve in practice. Sim-to-real RL, on the other hand, is often brittle due to the reality gap. This paper introduces SLAC, a method that renders real-world RL feasible for complex embodiments by leveraging a low-fidelity simulator to pretrain a task-agnostic latent action space. SLAC trains this latent action space via a customized unsupervised skill discovery method designed to promote temporal abstraction, disentanglement, and safety, thereby facilitating efficient downstream learning. Once a latent action space is learned, SLAC uses it as the action interface for a novel off-policy RL algorithm to autonomously learn downstream tasks through real-world interactions. We evaluate SLAC against existing methods on a suite of bimanual mobile manipulation tasks, where it achieves state-of-the-art performance. Notably, SLAC learns contact-rich whole-body tasks in under an hour of real-world interactions, without relying on any demonstrations or hand-crafted behavior priors. More information, code, and videos at robo-rl.github.io | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:41:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hu",
"Jiaheng",
""
],
[
"Stone",
"Peter",
""
],
[
"Martín-Martín",
"Roberto",
""
]
] |
2506.04148 | Michel Destrade | Sairam Pamulaparthi Venkata, Valentina Balbi, Michel Destradea, Dino Accoto, Giuseppe Zurlo | Programmable wrinkling for functionally-graded auxetic circular membranes | null | Extreme Mechanics Letters. Vol. 63 (2023) 102045 | 10.1016/j.eml.2023.102045 | null | cond-mat.soft | http://creativecommons.org/licenses/by/4.0/ | Materials with negative Poisson's ratio, also known as auxetic materials, display exotic properties such as expansion in all directions under uni-axial tension. For their unique properties, these materials find a broad range of applications in robotic, structural, aerospace, and biomedical engineering.
In this work we study the wrinkling behavior of thin and soft auxetic membranes, subjected to edge tractions. We show that spatial inhomogeneities of the Young modulus and of the Poisson ratio can be suitably tailored to produce non-trivial wrinkling patterns, with wrinkled regions that can appear, broaden, merge, and eventually disappear again, as the magnitude of applied tractions is increased monotonically. To model wrinkling in a functionally graded membrane, we employ the mathematically elegant and physically transparent tension field theory, an approximated method that we implement in commercially available software.
Beyond unveiling the challenging technological potential to achieve non-standard wrinkling on-demand in auxetic membranes, our study also confirms the potential of using tension field theory to study, analytically and numerically, instabilities in functionally graded materials. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:42:36 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Venkata",
"Sairam Pamulaparthi",
""
],
[
"Balbi",
"Valentina",
""
],
[
"Destradea",
"Michel",
""
],
[
"Accoto",
"Dino",
""
],
[
"Zurlo",
"Giuseppe",
""
]
] |
2506.04149 | Akira Taniguchi | Sugumi Kanno, Jiro Soda, and Akira Taniguchi | Parity violation in photon quasinormal modes of black holes | 22 pages, 1 figure | null | null | KOBE-COSMO-25-09 | gr-qc hep-th | http://creativecommons.org/publicdomain/zero/1.0/ | Given that black holes are ubiquitous in the universe and axion-like scalar fields are potential candidates for dark energy and/or dark matter, it is natural to consider cosmological black holes endowed with axion hair. We investigate the photon quasinormal modes of a Schwarzschild black hole with axion hair where the electromagnetic field is coupled to the axion field via a Chern-Simons interaction. We derive the master equations for the electromagnetic field as a set of coupled equations for parity-even and parity-odd modes and numerically compute quasinormal modes by using Leaver's continued fraction method. We find parity violation in the polarization of photons within the quasinormal mode spectrum. This parity violation in electromagnetic signals could serve as a new probe to explore the nature of the dark sector. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:42:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Kanno",
"Sugumi",
""
],
[
"Soda",
"Jiro",
""
],
[
"Taniguchi",
"Akira",
""
]
] |
2506.04150 | Eckhard Meinrenken | Eckhard Meinrenken | Introduction to moduli spaces and Dirac geometry | 50 pages | null | null | null | math.DG | http://creativecommons.org/licenses/by/4.0/ | Let $G$ be a Lie group, with an invariant metric on its Lie algebra $\mathfrak{g}$. Given a surface $\Sigma$ with boundary, and a collection of base points $\mathcal{V}\subset \Sigma$ meeting every boundary component, the moduli space (representation variety) $\mathcal{M}_G(\Sigma,\mathcal{V})$ carries a distinguished `quasi-symplectic' 2-form. We shall explain the finite-dimensional construction of this 2-form and discuss its basic properties, using quasi-Hamiltonian techniques and Dirac geometry. This article is an extended version of lectures given at the summer school 'Poisson 2024' at the Accademia Pontaniana in Napoli, July 2024. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:43:09 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Meinrenken",
"Eckhard",
""
]
] |
2506.04151 | Boris Post | Jan de Boer, Joshua Kames-King and Boris Post | Surgery and statistics in 3d gravity | 21 pages + appendices, 10 figures | null | null | null | hep-th | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We extend the correspondence between universal statistical features of large-$c$ 2d CFTs and surgery methods in pure AdS$_3$ quantum gravity. In particular, we introduce a method that we call RMT surgery, which relates a large class of off-shell partition functions in 3d gravity to the spectral statistics of general CFT observables. We apply this method to construct and compute an off-shell Euclidean wormhole whose boundaries are four-punctured spheres, which captures level repulsion in the high-energy sector of the boundary CFT. Using a similar gluing prescription, we also explore a new class of off-shell torus wormholes with trumpet boundaries, contributing to statistical moments of the density of primary states. Lastly, we demonstrate that surgery methods can be used as an intermediate step towards computing Seifert manifolds directly in 3d gravity. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:44:40 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"de Boer",
"Jan",
""
],
[
"Kames-King",
"Joshua",
""
],
[
"Post",
"Boris",
""
]
] |
2506.04152 | Ryan Langman | Ryan Langman, Xuesong Yang, Paarth Neekhara, Shehzeen Hussain, Edresson Casanova, Evelina Bakhturina, Jason Li | HiFiTTS-2: A Large-Scale High Bandwidth Speech Dataset | Submitted to Interspeech 2025 | null | null | null | eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces HiFiTTS-2, a large-scale speech dataset designed for high-bandwidth speech synthesis. The dataset is derived from LibriVox audiobooks, and contains approximately 36.7k hours of English speech for 22.05 kHz training, and 31.7k hours for 44.1 kHz training. We present our data processing pipeline, including bandwidth estimation, segmentation, text preprocessing, and multi-speaker detection. The dataset is accompanied by detailed utterance and audiobook metadata generated by our pipeline, enabling researchers to apply data quality filters to adapt the dataset to various use cases. Experimental results demonstrate that our data pipeline and resulting dataset can facilitate the training of high-quality, zero-shot text-to-speech (TTS) models at high bandwidths. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:49:41 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Langman",
"Ryan",
""
],
[
"Yang",
"Xuesong",
""
],
[
"Neekhara",
"Paarth",
""
],
[
"Hussain",
"Shehzeen",
""
],
[
"Casanova",
"Edresson",
""
],
[
"Bakhturina",
"Evelina",
""
],
[
"Li",
"Jason",
""
]
] |
2506.04153 | Jakob Oldenburg | Morris Brooks, Jakob Oldenburg, Diane Saint Aubin, Benjamin Schlein | Third Order Upper Bound for the Ground State Energy of the Dilute Bose Gas | null | null | null | null | math-ph math.MP | http://creativecommons.org/licenses/by/4.0/ | We consider a dilute Bose gas in the thermodynamic limit. We prove an upper bound for the ground state energy per unit volume, capturing the expected third order term, as predicted by Wu, Hugenholtz-Pines and Sawada. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:50:56 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Brooks",
"Morris",
""
],
[
"Oldenburg",
"Jakob",
""
],
[
"Aubin",
"Diane Saint",
""
],
[
"Schlein",
"Benjamin",
""
]
] |
2506.04154 | Armando W. Guti\'errez | Armando W. Guti\'errez, Olavi Nevanlinna | Metric functionals and weak convergence | null | null | null | null | math.FA math.MG | http://creativecommons.org/licenses/by/4.0/ | We introduce a notion of weak convergence in arbitrary metric spaces. Metric functionals are key in our analysis: weak convergence of sequences in a given metric space is tested against all the metric functionals defined on said space. When restricted to bounded sequences in normed linear spaces, we prove that our notion of weak convergence agrees with the standard one. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:51:28 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gutiérrez",
"Armando W.",
""
],
[
"Nevanlinna",
"Olavi",
""
]
] |
2506.04155 | Gregorio Staffieri | Gregorio Staffieri, Michele Fabrizio | Signatures of the Fermi surface reconstruction of a doped Mott insulator in a slab geometry | null | null | null | null | cond-mat.str-el cond-mat.mes-hall cond-mat.mtrl-sci | http://creativecommons.org/licenses/by/4.0/ | We investigate a hole-doped Mott insulator in a slab geometry using the dynamical cluster approximation. We show that the enhancement of the correlation strength at the surface results in the remarkable evolution of the layer-projected Fermi surface, which exhibits hole-like pockets in the superficial layers, but gradually evolves into a single electron-like surface in the innermost layers. We further analyze the behavior of the Friedel oscillations induced by the surface and identify distinct signatures of the Fermi surface reconstruction as function of hole-doping. In addition, we introduce a computationally tractable quantity that diagnoses the same Fermi surface variation by the concurrent breakdown of Luttinger's theorem. Both the latter quantity and the Friedel oscillations serve as reliable indicators of the change in Fermi surface topology, without the need for any periodization in momentum space. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:51:48 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Staffieri",
"Gregorio",
""
],
[
"Fabrizio",
"Michele",
""
]
] |
2506.04156 | Sarvesh Soni | Sarvesh Soni and Dina Demner-Fushman | A Dataset for Addressing Patient's Information Needs related to Clinical Course of Hospitalization | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Patients have distinct information needs about their hospitalization that can be addressed using clinical evidence from electronic health records (EHRs). While artificial intelligence (AI) systems show promise in meeting these needs, robust datasets are needed to evaluate the factual accuracy and relevance of AI-generated responses. To our knowledge, no existing dataset captures patient information needs in the context of their EHRs. We introduce ArchEHR-QA, an expert-annotated dataset based on real-world patient cases from intensive care unit and emergency department settings. The cases comprise questions posed by patients to public health forums, clinician-interpreted counterparts, relevant clinical note excerpts with sentence-level relevance annotations, and clinician-authored answers. To establish benchmarks for grounded EHR question answering (QA), we evaluated three open-weight large language models (LLMs)--Llama 4, Llama 3, and Mixtral--across three prompting strategies: generating (1) answers with citations to clinical note sentences, (2) answers before citations, and (3) answers from filtered citations. We assessed performance on two dimensions: Factuality (overlap between cited note sentences and ground truth) and Relevance (textual and semantic similarity between system and reference answers). The final dataset contains 134 patient cases. The answer-first prompting approach consistently performed best, with Llama 4 achieving the highest scores. Manual error analysis supported these findings and revealed common issues such as omitted key clinical evidence and contradictory or hallucinated content. Overall, ArchEHR-QA provides a strong benchmark for developing and evaluating patient-centered EHR QA systems, underscoring the need for further progress toward generating factual and relevant responses in clinical contexts. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:55:08 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Soni",
"Sarvesh",
""
],
[
"Demner-Fushman",
"Dina",
""
]
] |
2506.04157 | Andreas Burkhart | Andreas Burkhart, Nils Kohl, Barbara Wohlmuth and Jan Zawallich | A robust matrix-free approach for large-scale non-isothermal high-contrast viscosity Stokes flow on blended domains with applications to geophysics | 42 pages, 16 figures, 3 tables | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | We consider a compressible Stokes problem in the quasi-stationary case coupled with a time dependent advection-diffusion equation with special emphasis on high viscosity contrast geophysical mantle convection applications. In space, we use a P2-P1 Taylor--Hood element which is generated by a blending approach to account for the non-planar domain boundary without compromising the stencil data structure of uniformly refined elements. In time, we apply an operator splitting approach for the temperature equation combining the BDF2 method for diffusion and a particle method for advection, resulting in an overall second order scheme. Within each time step, a stationary Stokes problem with a high viscosity contrast has to be solved for which we propose a matrix-free, robust and scalable iterative solver based on Uzawa type block preconditioners, polynomial Chebyshev smoothers and a BFBT type Schur complement approximation. Our implementation is using a hybrid hierarchical grid approach allowing for massively parallel, high resolution Earth convection simulations. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:56:06 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Burkhart",
"Andreas",
""
],
[
"Kohl",
"Nils",
""
],
[
"Wohlmuth",
"Barbara",
""
],
[
"Zawallich",
"Jan",
""
]
] |
2506.04158 | Yujia Hu | Yujia Hu, Songhua Liu, Zhenxiong Tan, Xingyi Yang and Xinchao Wang | Image Editing As Programs with Diffusion Models | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While diffusion models have achieved remarkable success in text-to-image generation, they encounter significant challenges with instruction-driven image editing. Our research highlights a key challenge: these models particularly struggle with structurally inconsistent edits that involve substantial layout changes. To mitigate this gap, we introduce Image Editing As Programs (IEAP), a unified image editing framework built upon the Diffusion Transformer (DiT) architecture. At its core, IEAP approaches instructional editing through a reductionist lens, decomposing complex editing instructions into sequences of atomic operations. Each operation is implemented via a lightweight adapter sharing the same DiT backbone and is specialized for a specific type of edit. Programmed by a vision-language model (VLM)-based agent, these operations collaboratively support arbitrary and structurally inconsistent transformations. By modularizing and sequencing edits in this way, IEAP generalizes robustly across a wide range of editing tasks, from simple adjustments to substantial structural changes. Extensive experiments demonstrate that IEAP significantly outperforms state-of-the-art methods on standard benchmarks across various editing scenarios. In these evaluations, our framework delivers superior accuracy and semantic fidelity, particularly for complex, multi-step instructions. Codes are available at https://github.com/YujiaHu1109/IEAP. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:57:24 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hu",
"Yujia",
""
],
[
"Liu",
"Songhua",
""
],
[
"Tan",
"Zhenxiong",
""
],
[
"Yang",
"Xingyi",
""
],
[
"Wang",
"Xinchao",
""
]
] |
2506.04159 | Maxim Dolgushev | Maxim Dolgushev and Olivier B\'enichou | Fluctuations in the number of local minima in discrete-time fractional Brownian motion | 12 pages (main text + SM) | null | null | null | cond-mat.stat-mech | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The analysis of local minima in time series data and random landscapes is essential across numerous scientific disciplines, offering critical insights into system dynamics. Recently, Kundu, Majumdar, and Schehr derived the exact distribution of the number of local minima for a broad class of Markovian symmetric walks [Phys. Rev. E \textbf{110}, 024137 (2024)]; however, many real-world systems are non-Markovian, typically due to interactions with possibly hidden degrees of freedom. This work investigates the statistical properties of local minima in discrete-time samples of fractional Brownian motion (fBm), a non-Markovian Gaussian process with stationary increments, widely used to model complex, anomalous diffusion phenomena. We derive expressions for the variance of the number of local minima $m_N$ in an $N$-step discrete-time fBm. In contrast to the mean of $m_N$, which depends solely on nearest-neighbor correlations, the variance captures all long-range correlations inherent in fBm. Remarkably, we find that the variance exhibits two distinct scaling regimes: for Hurst exponent $H < 3/4$, $\mathrm{Var}(m_N) \propto N$, conforming to the central limit theorem (CLT); whereas for $H > 3/4$, $\mathrm{Var}(m_N) \propto N^{4H-2}$, resulting in the breakdown of the CLT. These findings are supported by numerical simulations and provide deeper insight into the interplay between memory effects and statistical fluctuations in non-Markovian processes. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 16:58:11 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dolgushev",
"Maxim",
""
],
[
"Bénichou",
"Olivier",
""
]
] |
2506.04160 | Daniel Slaughter | Niranjan Shivaram, Richard Thurston, Ali Belkacem, Thorsten Weber, Liang Z. Tan, and Daniel S. Slaughter | Interplay between ultrafast electronic and librational dynamics in liquid nitrobenzene probed with two-color four-wave mixing | null | null | null | null | physics.chem-ph physics.optics | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present an experimental and theoretical study of the interplay between ultrafast electron dynamics and librational dynamics in liquid nitrobenzene. A femtosecond ultraviolet pulse and two femtosecond near infrared pulses interact with nitrobenzene molecules, generating a four-wave mixing nonlinear signal that is measured in the Optical Kerr Effect geometry. The near infrared nonlinear signal is measured to be non-zero only at negative time delays, corresponding to the near infrared pulses arriving earlier than the ultraviolet pulse. We perform time-dependent Quantum Master Equation calculations, which include a classical libration model, to simulate the experiment. The simulations support the conclusion that the near infrared pulses launch librational motion, while simultaneously creating electronic coherences that result in a libration-modulated electronic nonlinear response. Furthermore, we conclude that the measured nonlinear optical signal corresponds to a non-parametric process that leaves the molecules in an excited electronic state. This work provides new insight into ultrafast nonlinear optical interactions in liquids and is an important step towards probing ultrafast electronic coherences in large molecules in the liquid phase. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:00:35 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Shivaram",
"Niranjan",
""
],
[
"Thurston",
"Richard",
""
],
[
"Belkacem",
"Ali",
""
],
[
"Weber",
"Thorsten",
""
],
[
"Tan",
"Liang Z.",
""
],
[
"Slaughter",
"Daniel S.",
""
]
] |
2506.04161 | Parsa Alian | Parsa Alian, Martin Tang, Ali Mesbah | VISCA: Inferring Component Abstractions for Automated End-to-End Testing | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Providing optimal contextual input presents a significant challenge for automated end-to-end (E2E) test generation using large language models (LLMs), a limitation that current approaches inadequately address. This paper introduces Visual-Semantic Component Abstractor (VISCA), a novel method that transforms webpages into a hierarchical, semantically rich component abstraction. VISCA starts by partitioning webpages into candidate segments utilizing a novel heuristic-based segmentation method. These candidate segments subsequently undergo classification and contextual information extraction via multimodal LLM-driven analysis, facilitating their abstraction into a predefined vocabulary of user interface (UI) components. This component-centric abstraction offers a more effective contextual basis than prior approaches, enabling more accurate feature inference and robust E2E test case generation. Our evaluations demonstrate that the test cases generated by VISCA achieve an average feature coverage of 92%, exceeding the performance of the state-of-the-art LLM-based E2E test generation method by 16%. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:00:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Alian",
"Parsa",
""
],
[
"Tang",
"Martin",
""
],
[
"Mesbah",
"Ali",
""
]
] |
2506.04162 | Elisa Fazzari | E. Fazzari, M.G. Dainotti, G. Montani, and A. Melchiorri | The effective running Hubble constant in SNe Ia as a marker for the dark energy nature | 31 pages, 8 figures. Prepared for submission to JHEAP | null | null | null | astro-ph.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a new method that reveal the nature of dark energy (DE) evolution. Specifically, the method consists of studying the evolving trend regarding the effective running Hubble constant: when it increases, it indicates a quintessence nature, and when it decreases, it reveals a phantom behavior. Within the framework of the dark energy models we analyze three parameterizations: the $w$CDM model, a reduced Chevallier-Polarski-Linder (CPL) model and a new theoretical model based on the possible creation of dark energy by the time-varying gravitational field of the expanding Universe. For each DE model, we construct a theoretical effective running Hubble constant, i.e. a function of the redshift, which highlights the difference between modified dynamics and the $\Lambda$CDM-one. Furthermore, these dark energy models are compared to the phenomenological model of a decreasing trend of the Hubble constant as a function of the redshift, called the power-law model (PL) and the $\Lambda$CDM one. These three theoretical functions are fitted against the binned SNe Ia data samples, i.e. the Pantheon and the Master samples, the latter being a collection of SNe Ia from 4 catalogs: Dark Energy Survey (DES), PantheonPlus, Pantheon and Joint Lightcurve Analysis (JLA), without duplicated SNe Ia, called the Master sample. The main result of our study is that the phenomenological PL model is statistically favored compared to the other proposed scenarios, both for the Pantheon and the Master samples. At this stage, the SNe Ia data do not indicate that the evolution of dark energy models among the studied ones is favored respect to the $\Lambda$CDM. Nevertheless, the binned Pantheon sample allows for a discrimination of the nature of dark energy at least at the $1\,\sigma$ level via the fit of the effective running Hubble constant. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:01:45 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Fazzari",
"E.",
""
],
[
"Dainotti",
"M. G.",
""
],
[
"Montani",
"G.",
""
],
[
"Melchiorri",
"A.",
""
]
] |
2506.04163 | Ming Xu | Yadong Jiao, Xiaoyan Cheng, Yuansheng Tang, Ming Xu | On the Synthetic Channels in Polar Codes over Binary-Input Discrete Memoryless Channels | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Polar codes introduced by Arikan in 2009 are the first code family achieving the capacity of binary-input discrete memoryless channels (BIDMCs) with low-complexity encoding and decoding. Identifying unreliable synthetic channels in polar code construction is crucial. Currently, because of the large size of the output alphabets of synthetic channels, there is no effective approach to evaluate their reliability, except in the case that the underlying channels are binary erasure channels. This paper defines equivalence and symmetry based on the likelihood ratio profile of BIDMCs and characterizes symmetric BIDMCs as random switching channels (RSCs) of binary symmetric channels. By converting the generation of synthetic channels in polar code construction into algebraic operations on underlying channels, some compact representations of RSCs for these synthetic channels are derived. Moreover, a lower bound for the average number of elements that possess the same likelihood ratio within the output alphabet of any synthetic channel generated in polar codes is also derived. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:02:04 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jiao",
"Yadong",
""
],
[
"Cheng",
"Xiaoyan",
""
],
[
"Tang",
"Yuansheng",
""
],
[
"Xu",
"Ming",
""
]
] |
2506.04164 | Shuo Huang | Shuo Huang, Chris Ormel, Simon Portegies Zwart, Eiichiro Kokubo, Tian Yi | A Resonant Beginning for the Solar System Terrestrial Planets | resubmitted to ApJ | null | null | null | astro-ph.EP physics.geo-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the past two decades, transit surveys have revealed a class of planets with thick atmospheres -- sub-Neptunes -- that must have completed their accretion in protoplanet disks. When planets form in the gaseous disk, the gravitational interaction with the disk gas drives their migration and results in the trapping of neighboring planets in mean motion resonances, though these resonances can later be broken when the damping effects of disk gas or planetesimals wane. It is widely accepted that the outer Solar System gas giant planets originally formed in a resonant chain, which was later disrupted by dynamical instabilities. Here, we explore whether the early formation of the terrestrial planets in a resonance chain (including Theia) can evolve to the present configuration. Using N-body simulations, we demonstrate that the giant planet instability would also have destabilized the terrestrial resonance chain, triggering moon-forming giant impacts in 20--50\% of our simulated systems, dependent on the initial resonance architecture. After the instability, the eccentricity and inclination of the simulated planets match their present-day values. Under the proposed scenario, the current period ratio of 3.05 between Mars and Venus -- devoid of any special significance in traditional late formation models -- naturally arises as a relic of the former resonance chain. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:02:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Huang",
"Shuo",
""
],
[
"Ormel",
"Chris",
""
],
[
"Zwart",
"Simon Portegies",
""
],
[
"Kokubo",
"Eiichiro",
""
],
[
"Yi",
"Tian",
""
]
] |
2506.04165 | Yashas Samaga | Yashas Samaga, Varun Yerram, Spandana Raj Babbula, Prateek Jain, Praneeth Netrapalli | Faster Approx. Top-K: Harnessing the Full Power of Two Stages | Rejected from MLSys 2025 | null | null | null | cs.LG cs.DS | http://creativecommons.org/licenses/by/4.0/ | We consider the Top-$K$ selection problem, which aims to identify the largest-$K$ elements from an array. Top-$K$ selection arises in many machine learning algorithms and often becomes a bottleneck on accelerators, which are optimized for dense matrix multiplications. To address this problem, \citet{chern2022tpuknnknearestneighbor} proposed a fast two-stage \textit{approximate} Top-$K$ algorithm: (i) partition the input array and select the top-$1$ element from each partition, (ii) sort this \textit{smaller subset} and return the top $K$ elements. In this paper, we consider a generalized version of this algorithm, where the first stage selects top-$K'$ elements, for some $1 \leq K' \leq K$, from each partition. Our contributions are as follows: (i) we derive an expression for the expected recall of this generalized algorithm and show that choosing $K' > 1$ with fewer partitions in the first stage reduces the input size to the second stage more effectively while maintaining the same expected recall as the original algorithm, (ii) we derive a bound on the expected recall for the original algorithm in \citet{chern2022tpuknnknearestneighbor} that is provably tighter by a factor of $2$ than the one in that paper, and (iii) we implement our algorithm on Cloud TPUv5e and achieve around an order of magnitude speedups over the original algorithm without sacrificing recall on real-world tasks. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:04:09 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Samaga",
"Yashas",
""
],
[
"Yerram",
"Varun",
""
],
[
"Babbula",
"Spandana Raj",
""
],
[
"Jain",
"Prateek",
""
],
[
"Netrapalli",
"Praneeth",
""
]
] |
2506.04166 | Kyuseong Choi | Caleb Chin, Aashish Khubchandani, Harshvardhan Maskara, Kyuseong Choi, Jacob Feitelberg, Albert Gong, Manit Paul, Tathagata Sadhukhan, Anish Agarwal, Raaz Dwivedi | N$^2$: A Unified Python Package and Test Bench for Nearest Neighbor-Based Matrix Completion | 21 pages, 6 figures | null | null | null | cs.LG stat.CO stat.ML | http://creativecommons.org/licenses/by/4.0/ | Nearest neighbor (NN) methods have re-emerged as competitive tools for matrix completion, offering strong empirical performance and recent theoretical guarantees, including entry-wise error bounds, confidence intervals, and minimax optimality. Despite their simplicity, recent work has shown that NN approaches are robust to a range of missingness patterns and effective across diverse applications. This paper introduces N$^2$, a unified Python package and testbed that consolidates a broad class of NN-based methods through a modular, extensible interface. Built for both researchers and practitioners, N$^2$ supports rapid experimentation and benchmarking. Using this framework, we introduce a new NN variant that achieves state-of-the-art results in several settings. We also release a benchmark suite of real-world datasets, from healthcare and recommender systems to causal inference and LLM evaluation, designed to stress-test matrix completion methods beyond synthetic scenarios. Our experiments demonstrate that while classical methods excel on idealized data, NN-based techniques consistently outperform them in real-world settings. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:04:34 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chin",
"Caleb",
""
],
[
"Khubchandani",
"Aashish",
""
],
[
"Maskara",
"Harshvardhan",
""
],
[
"Choi",
"Kyuseong",
""
],
[
"Feitelberg",
"Jacob",
""
],
[
"Gong",
"Albert",
""
],
[
"Paul",
"Manit",
""
],
[
"Sadhukhan",
"Tathagata",
""
],
[
"Agarwal",
"Anish",
""
],
[
"Dwivedi",
"Raaz",
""
]
] |
2506.04167 | Matthew Russell | Matthew Russell, Aman Shah, Giles Blaney, Judith Amores, Mary Czerwinski, Robert J.K. Jacob | Neural and Cognitive Impacts of AI: The Influence of Task Subjectivity on Human-LLM Collaboration | 15 pages, 12 figures | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | AI-based interactive assistants are advancing human-augmenting technology, yet their effects on users' mental and physiological states remain under-explored. We address this gap by analyzing how Copilot for Microsoft Word, a LLM-based assistant, impacts users. Using tasks ranging from objective (SAT reading comprehension) to subjective (personal reflection), and with measurements including fNIRS, Empatica E4, NASA-TLX, and questionnaires, we measure Copilot's effects on users. We also evaluate users' performance with and without Copilot across tasks. In objective tasks, participants reported a reduction of workload and an increase in enjoyment, which was paired with objective performance increases. Participants reported reduced workload and increased enjoyment with no change in performance in a creative poetry writing task. However, no benefits due to Copilot use were reported in a highly subjective self-reflection task. Although no physiological changes were recorded due to Copilot use, task-dependent differences in prefrontal cortex activation offer complementary insights into the cognitive processes associated with successful and unsuccessful human-AI collaboration. These findings suggest that AI assistants' effectiveness varies with task type-particularly showing decreased usefulness in tasks that engage episodic memory-and presents a brain-network based hypothesis of human-AI collaboration. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:04:48 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Russell",
"Matthew",
""
],
[
"Shah",
"Aman",
""
],
[
"Blaney",
"Giles",
""
],
[
"Amores",
"Judith",
""
],
[
"Czerwinski",
"Mary",
""
],
[
"Jacob",
"Robert J. K.",
""
]
] |
2506.04168 | Seohong Park | Seohong Park, Kevin Frans, Deepinder Mann, Benjamin Eysenbach, Aviral Kumar, Sergey Levine | Horizon Reduction Makes RL Scalable | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we study the scalability of offline reinforcement learning (RL) algorithms. In principle, a truly scalable offline RL algorithm should be able to solve any given problem, regardless of its complexity, given sufficient data, compute, and model capacity. We investigate if and how current offline RL algorithms match up to this promise on diverse, challenging, previously unsolved tasks, using datasets up to 1000x larger than typical offline RL datasets. We observe that despite scaling up data, many existing offline RL algorithms exhibit poor scaling behavior, saturating well below the maximum performance. We hypothesize that the horizon is the main cause behind the poor scaling of offline RL. We empirically verify this hypothesis through several analysis experiments, showing that long horizons indeed present a fundamental barrier to scaling up offline RL. We then show that various horizon reduction techniques substantially enhance scalability on challenging tasks. Based on our insights, we also introduce a minimal yet scalable method named SHARSA that effectively reduces the horizon. SHARSA achieves the best asymptotic performance and scaling behavior among our evaluation methods, showing that explicitly reducing the horizon unlocks the scalability of offline RL. Code: https://github.com/seohongpark/horizon-reduction | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:06:54 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Park",
"Seohong",
""
],
[
"Frans",
"Kevin",
""
],
[
"Mann",
"Deepinder",
""
],
[
"Eysenbach",
"Benjamin",
""
],
[
"Kumar",
"Aviral",
""
],
[
"Levine",
"Sergey",
""
]
] |
2506.04169 | Levon Nurbekyan | Xu Wang and Samy Wu Fung and Levon Nurbekyan | A primal-dual price-optimization method for computing equilibrium prices in mean-field games models | null | null | null | null | math.OC cs.NA econ.TH math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a simple yet efficient Lagrangian method for computing equilibrium prices in a mean-field game price-formation model. We prove that equilibrium prices are optimal in terms of a suitable criterion and derive a primal-dual gradient-based algorithm for computing them. One of the highlights of our computational framework is the efficient, simple, and flexible implementation of the algorithm using modern automatic differentiation techniques. Our implementation is modular and admits a seamless extension to high-dimensional settings with more complex dynamics, costs, and equilibrium conditions. Additionally, automatic differentiation enables a versatile algorithm that requires only coding the cost functions of agents. It automatically handles the gradients of the costs, thereby eliminating the need to manually form the adjoint equations. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:07:22 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wang",
"Xu",
""
],
[
"Fung",
"Samy Wu",
""
],
[
"Nurbekyan",
"Levon",
""
]
] |
2506.04170 | Tomasz Stebel | Piotr Bia{\l}as, Piotr Korcyl, Tomasz Stebel and Dawid Zapolski | Estimation of the reduced density matrix and entanglement entropies using autoregressive networks | 9 pages, 7 figures | null | null | null | quant-ph cond-mat.stat-mech cs.LG hep-lat hep-th | http://creativecommons.org/licenses/by/4.0/ | We present an application of autoregressive neural networks to Monte Carlo simulations of quantum spin chains using the correspondence with classical two-dimensional spin systems. We use a hierarchy of neural networks capable of estimating conditional probabilities of consecutive spins to evaluate elements of reduced density matrices directly. Using the Ising chain as an example, we calculate the continuum limit of the ground state's von Neumann and R\'enyi bipartite entanglement entropies of an interval built of up to 5 spins. We demonstrate that our architecture is able to estimate all the needed matrix elements with just a single training for a fixed time discretization and lattice volume. Our method can be applied to other types of spin chains, possibly with defects, as well as to estimating entanglement entropies of thermal states at non-zero temperature. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:08:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Białas",
"Piotr",
""
],
[
"Korcyl",
"Piotr",
""
],
[
"Stebel",
"Tomasz",
""
],
[
"Zapolski",
"Dawid",
""
]
] |
2506.04171 | Utkarsh Utkarsh | Utkarsh Utkarsh, Pengfei Cai, Alan Edelman, Rafael Gomez-Bombarelli, Christopher Vincent Rackauckas | Physics-Constrained Flow Matching: Sampling Generative Models with Hard Constraints | 27 pages, 9 figures, 4 tables | null | null | null | cs.LG cs.AI cs.CE cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | Deep generative models have recently been applied to physical systems governed by partial differential equations (PDEs), offering scalable simulation and uncertainty-aware inference. However, enforcing physical constraints, such as conservation laws (linear and nonlinear) and physical consistencies, remains challenging. Existing methods often rely on soft penalties or architectural biases that fail to guarantee hard constraints. In this work, we propose Physics-Constrained Flow Matching (PCFM), a zero-shot inference framework that enforces arbitrary nonlinear constraints in pretrained flow-based generative models. PCFM continuously guides the sampling process through physics-based corrections applied to intermediate solution states, while remaining aligned with the learned flow and satisfying physical constraints. Empirically, PCFM outperforms both unconstrained and constrained baselines on a range of PDEs, including those with shocks, discontinuities, and sharp features, while ensuring exact constraint satisfaction at the final solution. Our method provides a general framework for enforcing hard constraints in both scientific and general-purpose generative models, especially in applications where constraint satisfaction is essential. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:12:37 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Utkarsh",
"Utkarsh",
""
],
[
"Cai",
"Pengfei",
""
],
[
"Edelman",
"Alan",
""
],
[
"Gomez-Bombarelli",
"Rafael",
""
],
[
"Rackauckas",
"Christopher Vincent",
""
]
] |
2506.04172 | Lydia Manikonda | Shreenidhi Srinivasan and Lydia Manikonda | Does Prompt Design Impact Quality of Data Imputation by LLMs? | 7 pages | null | null | null | cs.LG cs.ET | http://creativecommons.org/licenses/by/4.0/ | Generating realistic synthetic tabular data presents a critical challenge in machine learning. It adds another layer of complexity when this data contain class imbalance problems. This paper presents a novel token-aware data imputation method that leverages the in-context learning capabilities of large language models. This is achieved through the combination of a structured group-wise CSV-style prompting technique and the elimination of irrelevant contextual information in the input prompt. We test this approach with two class-imbalanced binary classification datasets and evaluate the effectiveness of imputation using classification-based evaluation metrics. The experimental results demonstrate that our approach significantly reduces the input prompt size while maintaining or improving imputation quality compared to our baseline prompt, especially for datasets that are of relatively smaller in size. The contributions of this presented work is two-fold -- 1) it sheds light on the importance of prompt design when leveraging LLMs for synthetic data generation and 2) it addresses a critical gap in LLM-based data imputation for class-imbalanced datasets with missing data by providing a practical solution within computational constraints. We hope that our work will foster further research and discussions about leveraging the incredible potential of LLMs and prompt engineering techniques for synthetic data generation. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:15:19 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Srinivasan",
"Shreenidhi",
""
],
[
"Manikonda",
"Lydia",
""
]
] |
2506.04173 | Savannah Hays | Savannah P. Hays and Lianrui Zuo and Anqi Feng and Yihao Liu and Blake E. Dewey and Jiachen Zhuo and Ellen M. Mowry and Scott D. Newsome Jerry L. Prince and Aaron Carass | Synthetic multi-inversion time magnetic resonance images for visualization of subcortical structures | Under review at the Journal of Medical Imaging | null | null | null | eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Purpose: Visualization of subcortical gray matter is essential in neuroscience and clinical practice, particularly for disease understanding and surgical planning.While multi-inversion time (multi-TI) T$_1$-weighted (T$_1$-w) magnetic resonance (MR) imaging improves visualization, it is rarely acquired in clinical settings. Approach: We present SyMTIC (Synthetic Multi-TI Contrasts), a deep learning method that generates synthetic multi-TI images using routinely acquired T$_1$-w, T$_2$-weighted (T$_2$-w), and FLAIR images. Our approach combines image translation via deep neural networks with imaging physics to estimate longitudinal relaxation time (T$_1$) and proton density (PD) maps. These maps are then used to compute multi-TI images with arbitrary inversion times. Results: SyMTIC was trained using paired MPRAGE and FGATIR images along with T$_2$-w and FLAIR images. It accurately synthesized multi-TI images from standard clinical inputs, achieving image quality comparable to that from explicitly acquired multi-TI data.The synthetic images, especially for TI values between 400-800 ms, enhanced visualization of subcortical structures and improved segmentation of thalamic nuclei. Conclusion: SyMTIC enables robust generation of high-quality multi-TI images from routine MR contrasts. It generalizes well to varied clinical datasets, including those with missing FLAIR images or unknown parameters, offering a practical solution for improving brain MR image visualization and analysis. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:17:21 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hays",
"Savannah P.",
""
],
[
"Zuo",
"Lianrui",
""
],
[
"Feng",
"Anqi",
""
],
[
"Liu",
"Yihao",
""
],
[
"Dewey",
"Blake E.",
""
],
[
"Zhuo",
"Jiachen",
""
],
[
"Mowry",
"Ellen M.",
""
],
[
"Prince",
"Scott D. Newsome Jerry L.",
""
],
[
"Carass",
"Aaron",
""
]
] |
2506.04174 | Hengyu Liu | Hengyu Liu and Yuehao Wang and Chenxin Li and Ruisi Cai and Kevin Wang and Wuyang Li and Pavlo Molchanov and Peihao Wang and Zhangyang Wang | FlexGS: Train Once, Deploy Everywhere with Many-in-One Flexible 3D Gaussian Splatting | CVPR 2025; Project Page: https://flexgs.github.io | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | 3D Gaussian splatting (3DGS) has enabled various applications in 3D scene representation and novel view synthesis due to its efficient rendering capabilities. However, 3DGS demands relatively significant GPU memory, limiting its use on devices with restricted computational resources. Previous approaches have focused on pruning less important Gaussians, effectively compressing 3DGS but often requiring a fine-tuning stage and lacking adaptability for the specific memory needs of different devices. In this work, we present an elastic inference method for 3DGS. Given an input for the desired model size, our method selects and transforms a subset of Gaussians, achieving substantial rendering performance without additional fine-tuning. We introduce a tiny learnable module that controls Gaussian selection based on the input percentage, along with a transformation module that adjusts the selected Gaussians to complement the performance of the reduced model. Comprehensive experiments on ZipNeRF, MipNeRF and Tanks\&Temples scenes demonstrate the effectiveness of our approach. Code is available at https://flexgs.github.io. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:17:57 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Liu",
"Hengyu",
""
],
[
"Wang",
"Yuehao",
""
],
[
"Li",
"Chenxin",
""
],
[
"Cai",
"Ruisi",
""
],
[
"Wang",
"Kevin",
""
],
[
"Li",
"Wuyang",
""
],
[
"Molchanov",
"Pavlo",
""
],
[
"Wang",
"Peihao",
""
],
[
"Wang",
"Zhangyang",
""
]
] |
2506.04175 | Dominic Bernardi | Dominic Bernardi, Yajie Yuan, Alexander Y. Chen | Global Kinetic Simulations of Monster Shocks and Their Emission | 8 pages with 4 figures and 2 pages of supplemental materials with 2 figures | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fast magnetosonic waves are one of the two low-frequency plasma modes that can exist in a neutron star magnetosphere. It was recently realized that these waves may become nonlinear within the magnetosphere and steepen into some of the strongest shocks in the universe. These shocks, when in the appropriate parameter regime, may emit GHz radiation in the form of precursor waves. We present the first global Particle-in-Cell simulations of the nonlinear steepening of fast magnetosonic waves in a dipolar magnetosphere, and quantitatively demonstrate the strong plasma acceleration in the upstream of these shocks. In these simulations, we observe the production of precursor waves in a finite angular range. Using analytic scaling relations, we predict the expected frequency, power, and duration of this emission. Within a reasonable range of progenitor wave parameters, these precursor waves can reproduce many aspects of FRB observations. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:18:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Bernardi",
"Dominic",
""
],
[
"Yuan",
"Yajie",
""
],
[
"Chen",
"Alexander Y.",
""
]
] |
2506.04176 | Sudarshan Kumar Kenettinkara | Nikhil Manoj, G. D. Veerappa Gowda and Sudarshan Kumar K | A MUSCL-Hancock scheme for non-local conservation laws | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this article, we propose a MUSCL-Hancock-type second-order scheme for the discretization of a general class of non-local conservation laws and present its convergence analysis. The main difficulty in designing a MUSCL-Hancock-type scheme for non-local equations lies in the discretization of the convolution term, which we carefully formulate to ensure second-order accuracy and facilitate rigorous convergence analysis. We derive several essential estimates including $\mathrm{L}^\infty,$ bounded variation ($\mathrm{BV}$) and $\mathrm{L}^1$- Lipschitz continuity in time, which together with the Kolmogorov's compactness theorem yield the convergence of the approximate solutions to a weak solution. Further, by incorporating a mesh-dependent modification in the slope limiter, we establish convergence to the entropy solution. Numerical experiments are provided to validate the theoretical results and to demonstrate the improved accuracy of the proposed scheme over its first-order counterpart. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:20:13 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Manoj",
"Nikhil",
""
],
[
"Gowda",
"G. D. Veerappa",
""
],
[
"K",
"Sudarshan Kumar",
""
]
] |
2506.04177 | Olivier Debarre | Olivier Debarre, Chen Jiang | Numerical invariants of hyper-K\"ahler manifolds | 12 pages with an appendix by Chen Jiang | null | null | null | math.AG | http://creativecommons.org/licenses/by/4.0/ | We study various constraints on the Beauville quadratic form and the Huybrechts-Riemann-Roch polynomial for hyper-K\"ahler manifolds, mostly in dimension 6 and in the presence of an isotropic class.
In an appendix, Chen Jiang proves that in general, the Huybrechts-Riemann-Roch polynomial can always be written as a linear combination with nonnegative coefficients of certain explicit polynomials with positive coefficients. This implies that the Huybrechts-Riemann-Roch polynomial satisfies a curious symmetry property | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:21:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Debarre",
"Olivier",
""
],
[
"Jiang",
"Chen",
""
]
] |
2506.04179 | Anhao Zhao | Anhao Zhao, Fanghua Ye, Yingqi Fan, Junlong Tong, Zhiwei Fei, Hui Su, and Xiaoyu Shen | SkipGPT: Dynamic Layer Pruning Reinvented with Token Awareness and Module Decoupling | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) achieve remarkable performance across tasks but incur substantial computational costs due to their deep, multi-layered architectures. Layer pruning has emerged as a strategy to alleviate these inefficiencies, but conventional static pruning methods overlook two critical dynamics inherent to LLM inference: (1) horizontal dynamics, where token-level heterogeneity demands context-aware pruning decisions, and (2) vertical dynamics, where the distinct functional roles of MLP and self-attention layers necessitate component-specific pruning policies. We introduce SkipGPT, a dynamic layer pruning framework designed to optimize computational resource allocation through two core innovations: (1) global token-aware routing to prioritize critical tokens, and (2) decoupled pruning policies for MLP and self-attention components. To mitigate training instability, we propose a two-stage optimization paradigm: first, a disentangled training phase that learns routing strategies via soft parameterization to avoid premature pruning decisions, followed by parameter-efficient LoRA fine-tuning to restore performance impacted by layer removal. Extensive experiments demonstrate that SkipGPT reduces over 40% of model parameters while matching or exceeding the performance of the original dense model across benchmarks. By harmonizing dynamic efficiency with preserved expressivity, SkipGPT advances the practical deployment of scalable, resource-aware LLMs. Our code is publicly available at: https://github.com/EIT-NLP/SkipGPT. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:26:31 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhao",
"Anhao",
""
],
[
"Ye",
"Fanghua",
""
],
[
"Fan",
"Yingqi",
""
],
[
"Tong",
"Junlong",
""
],
[
"Fei",
"Zhiwei",
""
],
[
"Su",
"Hui",
""
],
[
"Shen",
"Xiaoyu",
""
]
] |
2506.04180 | Yuhao Wu | Yuhao Wu, Yushi Bai, Zhiqiang Hu, Juanzi Li, Roy Ka-Wei Lee | SuperWriter: Reflection-Driven Long-Form Generation with Large Language Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Long-form text generation remains a significant challenge for large language models (LLMs), particularly in maintaining coherence, ensuring logical consistency, and preserving text quality as sequence length increases. To address these limitations, we propose SuperWriter-Agent, an agent-based framework designed to enhance the quality and consistency of long-form text generation. SuperWriter-Agent introduces explicit structured thinking-through planning and refinement stages into the generation pipeline, guiding the model to follow a more deliberate and cognitively grounded process akin to that of a professional writer. Based on this framework, we construct a supervised fine-tuning dataset to train a 7B SuperWriter-LM. We further develop a hierarchical Direct Preference Optimization (DPO) procedure that uses Monte Carlo Tree Search (MCTS) to propagate final quality assessments and optimize each generation step accordingly. Empirical results across diverse benchmarks demonstrate that SuperWriter-LM achieves state-of-the-art performance, surpassing even larger-scale baseline models in both automatic evaluation and human evaluation. Furthermore, comprehensive ablation studies demonstrate the effectiveness of hierarchical DPO and underscore the value of incorporating structured thinking steps to improve the quality of long-form text generation. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:27:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wu",
"Yuhao",
""
],
[
"Bai",
"Yushi",
""
],
[
"Hu",
"Zhiqiang",
""
],
[
"Li",
"Juanzi",
""
],
[
"Lee",
"Roy Ka-Wei",
""
]
] |
2506.04181 | Ivan Dutta | Ivan Dutta, Anamitra Mukherjee, Onur Erten, Kush Saha | Exactly solvable spin liquids in Kitaev bilayers and moir\'e superlattices | 15 pages, 12 figures | null | null | null | cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | Building on the recent advancements on moir\'e superlattices, we propose an exactly solvable model with Kitaev-type interactions on a bilayer honeycomb lattice for both AA stacking and moir\'e superlattices. Employing Monte Carlo simulations and variational analysis, we uncover a rich variety of phases where the intra and interlayer $\mathbb{Z}_2$ fluxes (visons) are arranged in a periodic fashion in the ground state, tuned by interlayer coupling and out-of-plane external magnetic field. We further extend our model to moir\'e superlattices at various commensurate twist angles around two distinct twist centers represented by $C_{3z}$ and $C_{6z}$ of the honeycomb lattice. Our simulations reveal generalized patterns of plaquette values correlated with the AA or AB stacking regions across the moir\'e unit cell. In addition, depending on the twist angle, twist center and interlayer coupling, moir\'e superlattices exhibit to a variety of gapped and gapless spin liquid phases and can also host corner and edge modes. Our results highlight the rich physics in bilayer and twisted bilayer models of exactly solvable quantum spin liquids. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:28:21 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Dutta",
"Ivan",
""
],
[
"Mukherjee",
"Anamitra",
""
],
[
"Erten",
"Onur",
""
],
[
"Saha",
"Kush",
""
]
] |
2506.04182 | Ruiqi Zhang | Ruiqi Zhang, Changyi Xiao, Yixin Cao | Long or short CoT? Investigating Instance-level Switch of Large Reasoning Models | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid advancement of large reasoning models, long Chain-of-Thought (CoT) prompting has demonstrated strong performance on complex tasks. However, this often comes with a significant increase in token usage. In this paper, we conduct a comprehensive empirical analysis comparing long and short CoT strategies. Our findings reveal that while long CoT can lead to performance improvements, its benefits are often marginal relative to its significantly higher token consumption. Specifically, long CoT tends to outperform when ample generation budgets are available, whereas short CoT is more effective under tighter budget constraints. These insights underscore the need for a dynamic approach that selects the proper CoT strategy based on task context and resource availability. To address this, we propose SwitchCoT, an automatic framework that adaptively chooses between long and short CoT strategies to balance reasoning accuracy and computational efficiency. Moreover, SwitchCoT is designed to be budget-aware, making it broadly applicable across scenarios with varying resource constraints. Experimental results demonstrate that SwitchCoT can reduce inference costs by up to 50% while maintaining high accuracy. Notably, under limited token budgets, it achieves performance comparable to, or even exceeding, that of using either long or short CoT alone. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:28:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhang",
"Ruiqi",
""
],
[
"Xiao",
"Changyi",
""
],
[
"Cao",
"Yixin",
""
]
] |
2506.04183 | Maximilian Schaller | Maximilian Schaller, Alberto Bemporad, Stephen Boyd | Learning Parametric Convex Functions | null | null | null | null | math.OC | http://creativecommons.org/licenses/by/4.0/ | A parametrized convex function depends on a variable and a parameter, and is convex in the variable for any valid value of the parameter. Such functions can be used to specify parametrized convex optimization problems, i.e., a convex optimization family, in domain specific languages for convex optimization. In this paper we address the problem of fitting a parametrized convex function that is compatible with disciplined programming, to some given data. This allows us to fit a function arising in a convex optimization formulation directly to observed or simulated data. We demonstrate our open-source implementation on several examples, ranging from illustrative to practical. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:28:45 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Schaller",
"Maximilian",
""
],
[
"Bemporad",
"Alberto",
""
],
[
"Boyd",
"Stephen",
""
]
] |
2506.04184 | Andrea Simongini | Andrea Simongini, F. Ragosta, I. Di Palma, S. Piranomonte | Core-collapse supernova parameter estimation with the upcoming Vera C. Rubin Observatory | Paper accepted for publication at A&A | null | 10.1051/0004-6361/202554740 | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The Vera Rubin Observatory's Legacy Survey of Space and Time (LSST) is expected to revolutionize time-domain optical astronomy as we know it. With its unprecedented depth, the LSST will survey the southern hemisphere sky, generating nearly 32 trillion observations over its nominal 10-year operation. Among these, approximately 10 million will be supernovae (SNe). These observations will uniquely characterize the SN population, enabling studies of known and rare SN types, detailed parameterization of their light curves, deep searches for new SN progenitor populations, the discovery of strongly lensed SNe, and the compilation of a large, well-characterized sample of superluminous SNe. We analyzed a sample of 22663 simulations of LSST light curves for core collapse SNe (CCSNe), modeled using the radiative transfer code STELLA. We analyzed this dataset with the software CASTOR, which enables the reconstruction of synthetic light curves and spectra via a machine learning technique that allows one to retrieve the complete parameter map of a SN. For each parameter we compared the observed and the true values, determining how LSST light curves alone will contribute to characterize the progenitor and the explosion. Our results indicate that LSST alone will not suffice for a comprehensive and precise characterization of progenitor properties and explosion parameters. The limited spectral coverage of LSST light curves (in most cases) does not allow for the accurate estimation of bolometric luminosity, and consequently, of the explosion energy and nickel yield. Additionally, the redshift-absorption degeneracy is difficult to resolve without supplementary information. These findings suggest that for the most interesting SNe, complementary follow-up observations using spectrographs and optical facilities (particularly in the infrared bands) will be essential for accurate parameter determination. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:29:04 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Simongini",
"Andrea",
""
],
[
"Ragosta",
"F.",
""
],
[
"Di Palma",
"I.",
""
],
[
"Piranomonte",
"S.",
""
]
] |
2506.04185 | Qingfei Zhao | Qingfei Zhao, Ruobing Wang, Dingling Xu, Daren Zha, Limin Liu | R-Search: Empowering LLM Reasoning with Search via Multi-Reward Reinforcement Learning | 16 pages, 3 figures | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have notably progressed in multi-step and long-chain reasoning. However, extending their reasoning capabilities to encompass deep interactions with search remains a non-trivial challenge, as models often fail to identify optimal reasoning-search interaction trajectories, resulting in suboptimal responses. We propose R-Search, a novel reinforcement learning framework for Reasoning-Search integration, designed to enable LLMs to autonomously execute multi-step reasoning with deep search interaction, and learn optimal reasoning search interaction trajectories via multi-reward signals, improving response quality in complex logic- and knowledge-intensive tasks. R-Search guides the LLM to dynamically decide when to retrieve or reason, while globally integrating key evidence to enhance deep knowledge interaction between reasoning and search. During RL training, R-Search provides multi-stage, multi-type rewards to jointly optimize the reasoning-search trajectory. Experiments on seven datasets show that R-Search outperforms advanced RAG baselines by up to 32.2% (in-domain) and 25.1% (out-of-domain). The code and data are available at https://github.com/QingFei1/R-Search. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:29:22 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhao",
"Qingfei",
""
],
[
"Wang",
"Ruobing",
""
],
[
"Xu",
"Dingling",
""
],
[
"Zha",
"Daren",
""
],
[
"Liu",
"Limin",
""
]
] |
2506.04186 | Maryam Hussaini | Maryam Hussaini, Liam Connor, Ralf M. Konietzka, Vikram Ravi, Jakob Faber, Kritti Sharma and Myles Sherman | A Correlation Between FRB Dispersion Measure and Foreground Large-Scale Structure | null | null | null | null | astro-ph.CO astro-ph.GA astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | The distribution of baryons in the Universe remains a fundamental open question in astronomy, and the dispersion measure (DM) of Fast Radio Bursts (FRBs) serves as a valuable tool for probing this cosmic gas. We investigate the impact of the foreground cosmic web on FRB DMs using 61 localized FRBs and public galaxy catalogs. We test for the large-scale structure's impact on cosmological DM using two methods. First, we searched for a correlation between galaxy number density along the line of sight and extragalactic DM, and found a statistically significant positive correlation ($p$ = $1.76 \times 10^{-5}$). The shape of this correlation contains information about the cosmic baryon distribution, and can also be used to better constrain host galaxy DM by providing an estimate of the cosmic contribution on a per-source basis. We observe similar correlations in a mock FRB survey based on the IllustrisTNG cosmological simulation, where the DM is dominated by filaments in the IGM and not by halos. Next, we performed a stacking analysis that measures the average excess DM as a function of impact parameter of foreground galaxies to obtain spatial information about how ionized gas is distributed around galaxy halos. We report excess DM in the stacked signal for impact parameters up to Mpc scales ($\sim$3$\sigma$). Finally, we identified FRBs that do not appear to intersect intervening halos within $r_{vir}$, allowing us to estimate the fraction of baryons that reside in the IGM. We find $f_{\mathrm{IGM}} \geq 0.69$ at 95$\%$ confidence, indicating significant astrophysical feedback. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:31:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Hussaini",
"Maryam",
""
],
[
"Connor",
"Liam",
""
],
[
"Konietzka",
"Ralf M.",
""
],
[
"Ravi",
"Vikram",
""
],
[
"Faber",
"Jakob",
""
],
[
"Sharma",
"Kritti",
""
],
[
"Sherman",
"Myles",
""
]
] |
2506.04187 | Felipe Ramirez | Gilbert Michaud, Felipe A. Ram\'irez | Toward Khintchine's theorem with a moving target: extra divergence or finitely centered target | 22 pages | null | null | null | math.NT | http://creativecommons.org/licenses/by/4.0/ | Sz{\"u}sz's inhomogeneous version (1958) of Khintchine's theorem (1924) gives conditions on $\psi:\mathbb{N}\to\mathbb{R}_{\geq 0}$ under which for almost every real number $\alpha$ there exist infinitely many rationals $p/q$ such that \begin{equation*}
\lvert\alpha - \frac{p+\gamma}{q}\rvert < \frac{\psi(q)}{q}, \end{equation*} where $\gamma\in\mathbb{R}$ is some fixed inhomogeneous parameter. It is often interpreted as a statement about visits of $q\alpha\,(\bmod 1)$ to a shrinking target centered around $\gamma\,(\bmod 1)$, viewed in $\mathbb{R}/\mathbb{Z}$. Hauke and the second author have conjectured that Sz{\"u}sz's result continues to hold if the target is allowed to move as well as shrink, that is, if the inhomogeneous parameter $\gamma$ is allowed to depend on the denominator $q$ of the approximating rational. We show that the conjecture holds under an ``extra divergence'' assumption on $\psi$. We also show that it holds when the inhomogeneous parameter's movement is constrained to a finite set. As a byproduct, we obtain a finite-colorings version of the inhomogeneous Khintchine theorem, giving rational approximations with monochromatic denominators. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:32:03 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Michaud",
"Gilbert",
""
],
[
"Ramírez",
"Felipe A.",
""
]
] |
2506.04188 | Nicola Guglielmi | Nicola Guglielmi and Ernst Hairer | A fast and memoryless numerical method for solving fractional differential equations | null | null | null | null | math.NA cs.NA math.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The numerical solution of implicit and stiff differential equations by implicit numerical integrators has been largely investigated and there exist many excellent efficient codes available in the scientific community, as Radau5 (based on a Runge-Kutta collocation method at Radau points) and Dassl, based on backward differentiation formulas, among the others. When solving fractional ordinary differential equations (ODEs), the derivative operator is replaced by a non-local one and the fractional ODE is reformulated as a Volterra integral equation, to which these codes cannot be directly applied.
This article is a follow-up of the article by the authors (Guglielmi and Hairer, SISC, 2025) for differential equations with distributed delays. The main idea is to approximate the fractional kernel $t^{\alpha -1}/ \Gamma (\alpha )$ ($\alpha >0$) by a sum of exponential functions or by a sum of exponential functions multiplied by a monomial, and then to transform the fractional integral (of convolution type) into a set of ordinary differential equations. The augmented system is typically stiff and thus requires the use of an implicit method. It can have a very large dimension and requires a special treatment of the arising linear systems.
The present work presents an algorithm for the construction of an approximation of the fractional kernel by a sum of exponential functions, and it shows how the arising linear systems in a stiff time integrator can be solved efficiently. It is explained how the code Radau5 can be used for solving fractional differential equations. Numerical experiments illustrate the accuracy and the efficiency of the proposed method. Driver examples are publicly available from the homepages of the authors. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:36:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Guglielmi",
"Nicola",
""
],
[
"Hairer",
"Ernst",
""
]
] |
2506.04189 | Zhifei Yan | Wenchong Chen, Xinbu Cheng and Zhifei Yan | Colour-biased Hamilton cycles in randomly perturbed graphs | 28 pages, 2 figures | null | null | null | math.CO | http://creativecommons.org/licenses/by/4.0/ | Given a graph $G$ and an $r$-edge-colouring $\chi$ on $E(G)$, a Hamilton cycle $H\subset G$ is said to have $t$ colour-bias if $H$ contains $n/r+t$ edges of the same colour in $\chi$. Freschi, Hyde, Lada and Treglown showed every $r$-coloured graph $G$ on $n$ vertices with $\delta(G)\geq(r+1)n/2r+t$ contains a Hamilton cycle $H$ with $\Omega(t)$ colour-bias, generalizing a result of Balogh, Csaba, Jing and Pluh\'{a}r. In 2022, Gishboliner, Krivelevich and Michaeli proved that the random graph $G(n,m)$ with $m\geq(1/2+\varepsilon)n\log n$ typically admits an $\Omega(n)$ colour biased Hamilton cycle in any $r$-colouring.
In this paper, we investigate colour-biased Hamilton cycles in randomly perturbed graphs. We show that for every $\alpha>0$, adding $m=O(n)$ random edges to a graph $G_\alpha$ with $\delta(G_\alpha)\geq \alpha n$ typically ensures a Hamilton cycle with $\Omega(n)$ colour bias in any $r$-colouring of $G_\alpha\cup G(n,m)$. Conversely, for certain $G_{\alpha}$, reducing the number of random edges to $m=o(n)$ may eliminate all colour biased Hamilton cycles of $G(n,m)\cup G$ in a certain colouring. In contrast, at the critical endpoint $\alpha=(r+1)/2r$, adding $m$ random edges typically results in a Hamilton cycle with $\Omega(m)$ colour-bias for any $1\ll m\leq n$. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:36:49 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chen",
"Wenchong",
""
],
[
"Cheng",
"Xinbu",
""
],
[
"Yan",
"Zhifei",
""
]
] |
2506.04190 | Yuxuan Cao | Yuxuan Cao, Jiarong Xu, Chen Zhao, Jiaan Wang, Carl Yang, Chunping Wang, Yang Yang | How to Use Graph Data in the Wild to Help Graph Anomaly Detection? | Accepted by SIGKDD2025 | null | 10.1145/3690624.3709320 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In recent years, graph anomaly detection has found extensive applications in various domains such as social, financial, and communication networks. However, anomalies in graph-structured data present unique challenges, including label scarcity, ill-defined anomalies, and varying anomaly types, making supervised or semi-supervised methods unreliable. Researchers often adopt unsupervised approaches to address these challenges, assuming that anomalies deviate significantly from the normal data distribution. Yet, when the available data is insufficient, capturing the normal distribution accurately and comprehensively becomes difficult. To overcome this limitation, we propose to utilize external graph data (i.e., graph data in the wild) to help anomaly detection tasks. This naturally raises the question: How can we use external data to help graph anomaly detection tasks? To answer this question, we propose a framework called Wild-GAD. It is built upon a unified database, UniWildGraph, which comprises a large and diverse collection of graph data with broad domain coverage, ample data volume, and a unified feature space. Further, we develop selection criteria based on representativity and diversity to identify the most suitable external data for anomaly detection task. Extensive experiments on six real-world datasets demonstrate the effectiveness of Wild-GAD. Compared to the baseline methods, our framework has an average 18% AUCROC and 32% AUCPR improvement over the best-competing methods. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:37:47 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cao",
"Yuxuan",
""
],
[
"Xu",
"Jiarong",
""
],
[
"Zhao",
"Chen",
""
],
[
"Wang",
"Jiaan",
""
],
[
"Yang",
"Carl",
""
],
[
"Wang",
"Chunping",
""
],
[
"Yang",
"Yang",
""
]
] |
2506.04191 | Guillermo Vera De Salas | Ra\'ul Felipe and Guillermo Vera de Salas | Associative triple trisystems and standard embeddings | null | null | null | null | math.RA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Building on the established theories of Jordan triple disystems and Leibniz triple systems, we introduce and develop the theory of associative triple trisystems, filling a significant gap in the existing framework. We establish the classical relationships between associative, Jordan, and Lie triple systems within the context of trisystems. We present a significant example by equipping the space of matrices with a non-trivial associative dialgebra structure. We conclude defining the concept of di-endomorphisms of any module, which enables the construction of the standard embedding for any associative triple trisystem. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:38:18 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Felipe",
"Raúl",
""
],
[
"de Salas",
"Guillermo Vera",
""
]
] |
2506.04192 | Maria-Eleni Sfyraki | Maria-Eleni Sfyraki, Jun-Kun Wang | Lions and Muons: Optimization via Stochastic Frank-Wolfe | null | null | null | null | math.OC stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Stochastic Frank-Wolfe is a classical optimization method for solving constrained optimization problems. On the other hand, recent optimizers such as Lion and Muon have gained quite significant popularity in deep learning. In this work, we provide a unifying perspective by interpreting these seemingly disparate methods through the lens of Stochastic Frank-Wolfe. Specifically, we show that Lion and Muon with weight decay can be viewed as special instances of a Stochastic Frank-Wolfe, and we establish their convergence guarantees in terms of the Frank-Wolfe gap, a standard stationarity measure in non-convex optimization for Frank-Wolfe methods. We further find that convergence to this gap implies convergence to a KKT point of the original problem under a norm constraint for Lion and Muon. Moreover, motivated by recent empirical findings that stochastic gradients in modern machine learning tasks often exhibit heavy-tailed distributions, we extend Stochastic Frank-Wolfe to settings with heavy-tailed noise by developing two robust variants with strong theoretical guarantees, which in turn yields new variants of Lion and Muon. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:39:03 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Sfyraki",
"Maria-Eleni",
""
],
[
"Wang",
"Jun-Kun",
""
]
] |
2506.04193 | Stephen Pfohl | Stephen R. Pfohl, Natalie Harris, Chirag Nagpal, David Madras, Vishwali Mhasawade, Olawale Salaudeen, Awa Dieng, Shannon Sequeira, Santiago Arciniegas, Lillian Sung, Nnamdi Ezeanochie, Heather Cole-Lewis, Katherine Heller, Sanmi Koyejo, Alexander D'Amour | Understanding challenges to the interpretation of disaggregated evaluations of algorithmic fairness | null | null | null | null | stat.ML cs.CY cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Disaggregated evaluation across subgroups is critical for assessing the fairness of machine learning models, but its uncritical use can mislead practitioners. We show that equal performance across subgroups is an unreliable measure of fairness when data are representative of the relevant populations but reflective of real-world disparities. Furthermore, when data are not representative due to selection bias, both disaggregated evaluation and alternative approaches based on conditional independence testing may be invalid without explicit assumptions regarding the bias mechanism. We use causal graphical models to predict metric stability across subgroups under different data generating processes. Our framework suggests complementing disaggregated evaluations with explicit causal assumptions and analysis to control for confounding and distribution shift, including conditional independence testing and weighted performance estimation. These findings have broad implications for how practitioners design and interpret model assessments given the ubiquity of disaggregated evaluation. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:40:31 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Pfohl",
"Stephen R.",
""
],
[
"Harris",
"Natalie",
""
],
[
"Nagpal",
"Chirag",
""
],
[
"Madras",
"David",
""
],
[
"Mhasawade",
"Vishwali",
""
],
[
"Salaudeen",
"Olawale",
""
],
[
"Dieng",
"Awa",
""
],
[
"Sequeira",
"Shannon",
""
],
[
"Arciniegas",
"Santiago",
""
],
[
"Sung",
"Lillian",
""
],
[
"Ezeanochie",
"Nnamdi",
""
],
[
"Cole-Lewis",
"Heather",
""
],
[
"Heller",
"Katherine",
""
],
[
"Koyejo",
"Sanmi",
""
],
[
"D'Amour",
"Alexander",
""
]
] |
2506.04194 | Anay Mehrotra | Yang Cai and Alkis Kalavasis and Katerina Mamali and Anay Mehrotra and Manolis Zampetakis | What Makes Treatment Effects Identifiable? Characterizations and Estimators Beyond Unconfoundedness | Accepted for presentation at the 38th Conference on Learning Theory (COLT) 2025 | null | null | null | math.ST cs.LG econ.EM stat.ME stat.ML stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most of the widely used estimators of the average treatment effect (ATE) in causal inference rely on the assumptions of unconfoundedness and overlap. Unconfoundedness requires that the observed covariates account for all correlations between the outcome and treatment. Overlap requires the existence of randomness in treatment decisions for all individuals. Nevertheless, many types of studies frequently violate unconfoundedness or overlap, for instance, observational studies with deterministic treatment decisions -- popularly known as Regression Discontinuity designs -- violate overlap.
In this paper, we initiate the study of general conditions that enable the identification of the average treatment effect, extending beyond unconfoundedness and overlap. In particular, following the paradigm of statistical learning theory, we provide an interpretable condition that is sufficient and nearly necessary for the identification of ATE. Moreover, this condition characterizes the identification of the average treatment effect on the treated (ATT) and can be used to characterize other treatment effects as well. To illustrate the utility of our condition, we present several well-studied scenarios where our condition is satisfied and, hence, we prove that ATE can be identified in regimes that prior works could not capture. For example, under mild assumptions on the data distributions, this holds for the models proposed by Tan (2006) and Rosenbaum (2002), and the Regression Discontinuity design model introduced by Thistlethwaite and Campbell (1960). For each of these scenarios, we also show that, under natural additional assumptions, ATE can be estimated from finite samples.
We believe these findings open new avenues for bridging learning-theoretic insights and causal inference methodologies, particularly in observational studies with complex treatment mechanisms. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:40:55 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cai",
"Yang",
""
],
[
"Kalavasis",
"Alkis",
""
],
[
"Mamali",
"Katerina",
""
],
[
"Mehrotra",
"Anay",
""
],
[
"Zampetakis",
"Manolis",
""
]
] |
2506.04195 | Elena Zamaraeva | Elena Zamaraeva, Christopher M. Collins, George R. Darling, Matthew S. Dyer, Bei Peng, Rahul Savani, Dmytro Antypov, Vladimir V. Gusev, Judith Clymo, Paul G. Spirakis, Matthew J. Rosseinsky | MACS: Multi-Agent Reinforcement Learning for Optimization of Crystal Structures | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Geometry optimization of atomic structures is a common and crucial task in computational chemistry and materials design. Following the learning to optimize paradigm, we propose a new multi-agent reinforcement learning method called Multi-Agent Crystal Structure optimization (MACS) to address periodic crystal structure optimization. MACS treats geometry optimization as a partially observable Markov game in which atoms are agents that adjust their positions to collectively discover a stable configuration. We train MACS across various compositions of reported crystalline materials to obtain a policy that successfully optimizes structures from the training compositions as well as structures of larger sizes and unseen compositions, confirming its excellent scalability and zero-shot transferability. We benchmark our approach against a broad range of state-of-the-art optimization methods and demonstrate that MACS optimizes periodic crystal structures significantly faster, with fewer energy calculations, and the lowest failure rate. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:40:57 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zamaraeva",
"Elena",
""
],
[
"Collins",
"Christopher M.",
""
],
[
"Darling",
"George R.",
""
],
[
"Dyer",
"Matthew S.",
""
],
[
"Peng",
"Bei",
""
],
[
"Savani",
"Rahul",
""
],
[
"Antypov",
"Dmytro",
""
],
[
"Gusev",
"Vladimir V.",
""
],
[
"Clymo",
"Judith",
""
],
[
"Spirakis",
"Paul G.",
""
],
[
"Rosseinsky",
"Matthew J.",
""
]
] |
2506.04196 | Yuzhe Li | Yu-Zhe Li, Wen-Long Xu, Yi-Gu Chen and Wei-Hua Lei | Lense-Thirring Precession Modulates Repeated Lensing of Continues Gravitational Wave Source from AGN Disks | null | null | null | null | astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gravitational lensing of gravitational waves (GWs) offers a novel observational channel that complements traditional electromagnetic approaches and provides unique insights into the astrophysical environments of GW sources. In this work, we investigate the repeated lensing of continuous gravitational wave (CW) sources in active galactic nucleus (AGN) disks by central supermassive black holes (SMBHs), focusing on the imprint of SMBH spin via the Lense-Thirring (LT) effect. Although typically weak and challenging to observe, the spin-induced precession of source orbits can accumulate over time, thereby modulating the lensing geometry. Such modulations influence the magnification, duration, and waveform structure of each repeated lensing event, and enhance the overall probability of lensing occurrences. Using matched filtering, we demonstrate that spin-dependent signatures may be detectable, suggesting that lensed CW signals could serve as an indirect probe of SMBH spin in AGNs. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:41:56 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Yu-Zhe",
""
],
[
"Xu",
"Wen-Long",
""
],
[
"Chen",
"Yi-Gu",
""
],
[
"Lei",
"Wei-Hua",
""
]
] |
2506.04197 | Peixue Wu | Roy Araiza, Marius Junge, Peixue Wu | Transportation cost and contraction coefficient for channels on von Neumann algebras | null | null | null | null | math.OA math-ph math.FA math.MP quant-ph | http://creativecommons.org/licenses/by/4.0/ | We present a noncommutative optimal transport framework for quantum channels acting on von Neumann algebras. Our central object is the Lipschitz cost measure, a transportation-inspired quantity that evaluates the minimal cost required to move between quantum states via a given channel. Accompanying this is the Lipschitz contraction coefficient, which captures how much the channel contracts the Wasserstein-type distance between states. We establish foundational properties of these quantities, including continuity, dual formulations, and behavior under composition and tensorization. Applications include recovery of several mathematical quantities including expected group word length and Carnot-Carath\'eodory distance, via transportation cost. Moreover, we show that if the Lipschitz contraction coefficient is strictly less than one, one can get entropy contraction and mixing time estimates for certain classes of non-symmetric channels. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:42:33 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Araiza",
"Roy",
""
],
[
"Junge",
"Marius",
""
],
[
"Wu",
"Peixue",
""
]
] |
2506.04198 | Dimitris Ntotsikas Mr | D. Ntotsikas and K.N. Gourgouliatos | Interlinking internal and external magnetic fields of relativistically rotating neutron stars | 11 pages, 11 figures | null | null | null | astro-ph.HE | http://creativecommons.org/licenses/by/4.0/ | This work presents a global solution for the internal and the external field of an axisymmetric rotating neutron star. It is shown that the twist of the internal field affects the external field, by increasing the number of open field lines and eventually the spin-down rate of the star. This effect is far more drastic if the toroidal field, and consequently the poloidal current flowing within the star, is allowed to populate the closed field lines of the magnetosphere, rather than if it remains confined in the star. We further remark that the internal field structure depends on the presence of a twisted magnetosphere: if the twist current is not allowed to flow in the magnetosphere it only occupied a narrow toroid at the interior of the star, whereas if the twist currents are allowed to flow in the magnetosphere the internal toroidal field may occupy a significant volume of the stellar interior. Strong magnetospheric currents may also impact the emission mechanisms, and lead to fluctuations in magnetar spin-down rates, moding and nulling of pulsars, a correlation between angular shear and twist, and the general morphology of the pulsar magnetic field leading to various observational manifestations. The magnetospheric toroidal fields may possibly dissipate, thus the system may switch from global twist to internal twist and consequently exhibit transient behavior. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:43:40 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ntotsikas",
"D.",
""
],
[
"Gourgouliatos",
"K. N.",
""
]
] |
2506.04199 | Aake Nordlund | {\AA}ke Nordlund | Abundance Effects from Protoplanetary Disk Outflows | 5 pages, 4 figures, to be submitted to A&A | null | null | null | astro-ph.EP astro-ph.SR | http://creativecommons.org/licenses/by/4.0/ | Systematic abundance differences that depend on the condensation temperatures of elements have been observed, in particular for stars similar to the Sun; solar twins and solar analogs. Similar differences have also recently been shown to exist between solar abundances and abundances of refractory elements in primitive chondrites. Numerous mechanisms have been proposed to account for these effects, including also differences observed in binary systems. Rather than relying on specific mechanisms, this paper aims to show that the observed effects are a natural and unavoidable outcome of the star formation process itself, in which the associated outflows (winds and jets) carry away material, with efficiency varying with condensation temperature. By using analysis based on modeling results and scaling laws, the trends and magnitudes of the effects are investigated, in three contexts: 1) with respect to differences between the Sun and solar twins, 2) with respect to differences between the Sun and CI-chondrites, and 3) with respect to differences between members of binaries. It is shown that protoplanetary disk outflows indeed are expected to have differential abundance effects, with trends and magnitudes consistent with observed abundance effects. The qualitative as well as semi-quantitative character of the effects are reproduced, in all three contexts. The results demonstrate that the observed systematic differences are likely results of the disk outflows associated with the accretion process. In contrast to mechanisms relying on the tiny mass of planets leaving an observable signature, outflows carry away masses similar to the entire mass of the star, thus much more easily resulting in differential effects with the magnitudes observed, without for example having to assume that the abundance differences are limited to the convection zones of the stars. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:44:14 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Nordlund",
"Åke",
""
]
] |
2506.04200 | Sing Lam Wong | Sing Lam Wong, Andrew C. Potter | Entanglement renormalization circuits for $2d$ Gaussian Fermion States | null | null | null | null | quant-ph cond-mat.str-el | http://creativecommons.org/licenses/by/4.0/ | The simulation of entangled ground-states of quantum materials remains challenging for classical computational methods in more than one spatial dimension, and is a prime target for quantum computational advantage. To this end, an important goal is to identify efficient quantum state preparation protocols that minimize the physical qubit number and circuit depth resources required to capture higher-dimensional quantum correlations. This work introduces a quantum circuit compression algorithm for Gaussian fermion states based on the multi-scale entanglement renormalization ansatz (MERA), which provides an exponential reduction in the circuit depth required to approximate highly-entangled ground-states relevant for quantum materials simulations. The algorithm, termed two-dimensional Gaussian MERA ($2d$ GMERA), extends MERA techniques to compress higher-dimensional Gaussian states. Through numerical simulations of the Haldane model on a honeycomb lattice, the method is shown to accurately capture area-law entangled states including topologically trivial insulators, Chern insulators, and critical Dirac semimetals. While Gaussian states alone are classically simulable, this approach establishes empirical upper bounds on quantum resources needed to prepare free fermion states that are adiabatically connected to correlated ground states, providing guidance for implementing these protocols on near-term quantum devices and offering a foundation for simulating more complex quantum materials. Finally, we develop a novel fermion-to-qubit encoding scheme, based on an expanding $2d$ topological order, that enables implementing fermionic rotations via qubit Pauli rotations with constant Pauli weight independent of system size. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:44:17 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wong",
"Sing Lam",
""
],
[
"Potter",
"Andrew C.",
""
]
] |
2506.04201 | Xianzhe Tang | Xianzhe TZ Tang, Simon Birrer, Anowar J. Shajib, Narayan Khadka, the LSST Strong Gravitational Lensing Science Collaboration, the LSST Dark Energy Science Collaboration | Analyzing Line-of-sight selection biases in galaxy-scale strong lensing with external convergence and shear | 27 pages, 8 figures, 2 tables | null | null | null | astro-ph.CO astro-ph.GA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The upcoming Vera Rubin Observatory Legacy Survey of Space and Time (LSST) will dramatically increase the number of strong gravitational lensing systems, requiring precise modeling of line-of-sight (LOS) effects to mitigate biases in lensing observations and cosmological inferences. We develop a method to construct joint distributions of external convergence ($\kappa_{\mathrm{ext}}$) and shear ($\gamma_{\mathrm{ext}}$) for strong lensing LOS by aggregating large-scale structure simulations with high-resolution halo renderings and non-linear correction. Our approach captures both smooth background matter and perturbations from halos, enabling accurate modeling of LOS effects. We apply non-linear LOS corrections to $\kappa_{\mathrm{ext}}$ and $\gamma_{\mathrm{ext}}$ that address the non-additive lensing effects caused by objects along the LOS in strong lensing. We find that, with a minimum image separation of $1.0^{\prime\prime}$, non-linear LOS correction due to the presence of a dominant deflector slightly increases the ratio of quadruple to double lenses; this non-linear LOS correction also introduces systematic biases of $\sim 0.1\%$ for galaxy-AGN lenses in the inferred Hubble constant ($H_0$) if not accounted for. We also observe a $0.66\%$ bias for galaxy-galaxy lenses on $H_0$, and even larger biases up to $1.02\%$ for galaxy-AGN systems if LOS effects are not accounted for. These results highlight the importance of LOS for precision cosmology. The publicly available code and datasets provide tools for incorporating LOS effects in future analyses. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:46:02 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Tang",
"Xianzhe TZ",
""
],
[
"Birrer",
"Simon",
""
],
[
"Shajib",
"Anowar J.",
""
],
[
"Khadka",
"Narayan",
""
],
[
"Collaboration",
"the LSST Strong Gravitational Lensing Science",
""
],
[
"Collaboration",
"the LSST Dark Energy Science",
""
]
] |
2506.04202 | Yanting Wang | Yanting Wang, Wei Zou, Runpeng Geng, Jinyuan Jia | TracLLM: A Generic Framework for Attributing Long Context LLMs | To appear in USENIX Security Symposium 2025. The code and data are at: https://github.com/Wang-Yanting/TracLLM | null | null | null | cs.CR cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Long context large language models (LLMs) are deployed in many real-world applications such as RAG, agent, and broad LLM-integrated applications. Given an instruction and a long context (e.g., documents, PDF files, webpages), a long context LLM can generate an output grounded in the provided context, aiming to provide more accurate, up-to-date, and verifiable outputs while reducing hallucinations and unsupported claims. This raises a research question: how to pinpoint the texts (e.g., sentences, passages, or paragraphs) in the context that contribute most to or are responsible for the generated output by an LLM? This process, which we call context traceback, has various real-world applications, such as 1) debugging LLM-based systems, 2) conducting post-attack forensic analysis for attacks (e.g., prompt injection attack, knowledge corruption attacks) to an LLM, and 3) highlighting knowledge sources to enhance the trust of users towards outputs generated by LLMs. When applied to context traceback for long context LLMs, existing feature attribution methods such as Shapley have sub-optimal performance and/or incur a large computational cost. In this work, we develop TracLLM, the first generic context traceback framework tailored to long context LLMs. Our framework can improve the effectiveness and efficiency of existing feature attribution methods. To improve the efficiency, we develop an informed search based algorithm in TracLLM. We also develop contribution score ensemble/denoising techniques to improve the accuracy of TracLLM. Our evaluation results show TracLLM can effectively identify texts in a long context that lead to the output of an LLM. Our code and data are at: https://github.com/Wang-Yanting/TracLLM. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:48:16 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wang",
"Yanting",
""
],
[
"Zou",
"Wei",
""
],
[
"Geng",
"Runpeng",
""
],
[
"Jia",
"Jinyuan",
""
]
] |
2506.04203 | Youhe Jiang | Youhe Jiang, Fangcheng Fu, Wanru Zhao, Stephan Rabanser, Nicholas D. Lane, Binhang Yuan | Cascadia: A Cascade Serving System for Large Language Models | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in large language models (LLMs) have intensified the need to deliver both rapid responses and high-quality answers. More powerful models yield better results but incur higher inference latency, whereas smaller models are faster yet less capable. Recent work proposes balancing this latency-quality trade-off using model cascades, which route simpler queries to smaller models and more complex ones to larger models. However, enabling efficient cascade serving remains challenging. Current frameworks lack effective mechanisms for handling (i) the huge and varying resource demands of different LLMs, (ii) the inherent heterogeneity of LLM workloads, and (iii) the co-optimization of system deployment and routing strategy. Motivated by these observations, we introduce Cascadia, a novel cascade serving framework designed explicitly to schedule request routing and deploy model cascades for fast, quality-preserving LLM serving. Cascadia employs a bi-level optimization method: at the inner level, it uses a mixed-integer linear program to select resource allocations and parallelism strategies based on LLM information and workload characteristics; at the outer level, it applies a weighted Tchebycheff algorithm to iteratively co-optimize the routing strategy and the system deployment produced by the inner level. Our extensive evaluation on diverse workload traces and different model cascades (DeepSeek and the Llama series) demonstrates that Cascadia significantly outperforms both single-model deployments and the state-of-the-art cascade serving baseline, achieving up to 4x (2.3x on average) tighter latency SLOs and up to 5x (2.4x on average) higher throughput while maintaining target answer quality. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:48:38 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jiang",
"Youhe",
""
],
[
"Fu",
"Fangcheng",
""
],
[
"Zhao",
"Wanru",
""
],
[
"Rabanser",
"Stephan",
""
],
[
"Lane",
"Nicholas D.",
""
],
[
"Yuan",
"Binhang",
""
]
] |
2506.04204 | Martin Beseda | Martin Beseda, Vittorio Cortellessa, Daniele Di Pompeo, Luca Traini, Michele Tucci | A Kernel-Based Approach for Accurate Steady-State Detection in Performance Time Series | This manuscript is under review by Future Generation Computer Systems | null | null | null | cs.PF cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | This paper addresses the challenge of accurately detecting the transition from the warmup phase to the steady state in performance metric time series, which is a critical step for effective benchmarking. The goal is to introduce a method that avoids premature or delayed detection, which can lead to inaccurate or inefficient performance analysis. The proposed approach adapts techniques from the chemical reactors domain, detecting steady states online through the combination of kernel-based step detection and statistical methods. By using a window-based approach, it provides detailed information and improves the accuracy of identifying phase transitions, even in noisy or irregular time series. Results show that the new approach reduces total error by 14.5% compared to the state-of-the-art method. It offers more reliable detection of the steady-state onset, delivering greater precision for benchmarking tasks. For users, the new approach enhances the accuracy and stability of performance benchmarking, efficiently handling diverse time series data. Its robustness and adaptability make it a valuable tool for real-world performance evaluation, ensuring consistent and reproducible results. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:48:42 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Beseda",
"Martin",
""
],
[
"Cortellessa",
"Vittorio",
""
],
[
"Di Pompeo",
"Daniele",
""
],
[
"Traini",
"Luca",
""
],
[
"Tucci",
"Michele",
""
]
] |
2506.04205 | Jinghan Jia | Jinghan Jia, Hadi Reisizadeh, Chongyu Fan, Nathalie Baracaldo, Mingyi Hong, Sijia Liu | EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have shown remarkable reasoning capabilities when trained with chain-of-thought (CoT) supervision. However, the long and verbose CoT traces, especially those distilled from large reasoning models (LRMs) such as DeepSeek-R1, significantly increase training costs during the distillation process, where a non-reasoning base model is taught to replicate the reasoning behavior of an LRM. In this work, we study the problem of CoT condensation for resource-efficient reasoning training, aimed at pruning intermediate reasoning steps (i.e., thoughts) in CoT traces, enabling supervised model training on length-reduced CoT data while preserving both answer accuracy and the model's ability to generate coherent reasoning. Our rationale is that CoT traces typically follow a three-stage structure: problem understanding, exploration, and solution convergence. Through empirical analysis, we find that retaining the structure of the reasoning trace, especially the early stage of problem understanding (rich in reflective cues) and the final stage of solution convergence, is sufficient to achieve lossless reasoning supervision. To this end, we propose an Edge-Preserving Condensation method, EPiC, which selectively retains only the initial and final segments of each CoT trace while discarding the middle portion. This design draws an analogy to preserving the "edge" of a reasoning trajectory, capturing both the initial problem framing and the final answer synthesis, to maintain logical continuity. Experiments across multiple model families (Qwen and LLaMA) and benchmarks show that EPiC reduces training time by over 34% while achieving lossless reasoning accuracy on MATH500, comparable to full CoT supervision. To the best of our knowledge, this is the first study to explore thought-level CoT condensation for efficient reasoning model distillation. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:49:10 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Jia",
"Jinghan",
""
],
[
"Reisizadeh",
"Hadi",
""
],
[
"Fan",
"Chongyu",
""
],
[
"Baracaldo",
"Nathalie",
""
],
[
"Hong",
"Mingyi",
""
],
[
"Liu",
"Sijia",
""
]
] |
2506.04206 | Reza Ramezanpour | Reza Ramezanpour, Victor M. Tenorio, Antonio G. Marques, Ashutosh Sabharwal, Santiago Segarra | A Few Moments Please: Scalable Graphon Learning via Moment Matching | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Graphons, as limit objects of dense graph sequences, play a central role in the statistical analysis of network data. However, existing graphon estimation methods often struggle with scalability to large networks and resolution-independent approximation, due to their reliance on estimating latent variables or costly metrics such as the Gromov-Wasserstein distance. In this work, we propose a novel, scalable graphon estimator that directly recovers the graphon via moment matching, leveraging implicit neural representations (INRs). Our approach avoids latent variable modeling by training an INR--mapping coordinates to graphon values--to match empirical subgraph counts (i.e., moments) from observed graphs. This direct estimation mechanism yields a polynomial-time solution and crucially sidesteps the combinatorial complexity of Gromov-Wasserstein optimization. Building on foundational results, we establish a theoretical guarantee: when the observed subgraph motifs sufficiently represent those of the true graphon (a condition met with sufficiently large or numerous graph samples), the estimated graphon achieves a provable upper bound in cut distance from the ground truth. Additionally, we introduce MomentMixup, a data augmentation technique that performs mixup in the moment space to enhance graphon-based learning. Our graphon estimation method achieves strong empirical performance--demonstrating high accuracy on small graphs and superior computational efficiency on large graphs--outperforming state-of-the-art scalable estimators in 75\% of benchmark settings and matching them in the remaining cases. Furthermore, MomentMixup demonstrated improved graph classification accuracy on the majority of our benchmarks. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:51:01 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ramezanpour",
"Reza",
""
],
[
"Tenorio",
"Victor M.",
""
],
[
"Marques",
"Antonio G.",
""
],
[
"Sabharwal",
"Ashutosh",
""
],
[
"Segarra",
"Santiago",
""
]
] |
2506.04207 | Shuang Chen Sky | Shuang Chen, Yue Guo, Zhaochen Su, Yafu Li, Yulun Wu, Jiacheng Chen, Jiayu Chen, Weijie Wang, Xiaoye Qu, Yu Cheng | Advancing Multimodal Reasoning: From Optimized Cold Start to Staged Reinforcement Learning | 19 pages, 6 figures | null | null | null | cs.LG cs.AI cs.CL cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Inspired by the remarkable reasoning capabilities of Deepseek-R1 in complex textual tasks, many works attempt to incentivize similar capabilities in Multimodal Large Language Models (MLLMs) by directly applying reinforcement learning (RL). However, they still struggle to activate complex reasoning. In this paper, rather than examining multimodal RL in isolation, we delve into current training pipelines and identify three crucial phenomena: 1) Effective cold start initialization is critical for enhancing MLLM reasoning. Intriguingly, we find that initializing with carefully selected text data alone can lead to performance surpassing many recent multimodal reasoning models, even before multimodal RL. 2) Standard GRPO applied to multimodal RL suffers from gradient stagnation, which degrades training stability and performance. 3) Subsequent text-only RL training, following the multimodal RL phase, further enhances multimodal reasoning. This staged training approach effectively balances perceptual grounding and cognitive reasoning development. By incorporating the above insights and addressing multimodal RL issues, we introduce ReVisual-R1, achieving a new state-of-the-art among open-source 7B MLLMs on challenging benchmarks including MathVerse, MathVision, WeMath, LogicVista, DynaMath, and challenging AIME2024 and AIME2025. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:51:08 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chen",
"Shuang",
""
],
[
"Guo",
"Yue",
""
],
[
"Su",
"Zhaochen",
""
],
[
"Li",
"Yafu",
""
],
[
"Wu",
"Yulun",
""
],
[
"Chen",
"Jiacheng",
""
],
[
"Chen",
"Jiayu",
""
],
[
"Wang",
"Weijie",
""
],
[
"Qu",
"Xiaoye",
""
],
[
"Cheng",
"Yu",
""
]
] |
2506.04208 | Haoran Guan Mr | Haoran Guan and Yuwei Fan | Rounding error analysis of randomized CholeskyQR2 for sparse matrices | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work focuses on rounding error analysis of randomized CholeskyQR2 (RCholeskyQR2) for sparse matrices. We form RCholeskyQR2 with CholeskyQR2 and matrix sketching in order to accelerate CholeskyQR2 and improve its applicability by compressing the input matrix $X \in \mathbb{R}^{m\times n}$. In many real-world applications, the input matrix $X$ is always sparse. In this work, we follow the model of sparse matrices proposed in \cite{CSparse} and provide an improved rounding error analysis of RCholeskyQR2 for sparse matrices. We define a new matrix norm $\norm{\cdot}_{g}$ and use its properties in rounding error analysis. $\norm{\cdot}_{F}$ and its connections with rounding error analysis is also utilized in CholeskyQR for the first time. Our analysis is an improvement compared to that in \cite{CSparse, error}. We provide the corresponding sufficient conditions for different types of sparse matrices based on the existence of dense columns, together with error bounds of both orthogonality and residual. Numerical experiments demonstrate our findings and show the effectiveness of matrix sketching on CholeskyQR2. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:51:39 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Guan",
"Haoran",
""
],
[
"Fan",
"Yuwei",
""
]
] |
2506.04209 | Jingfeng Yang | Jingfeng Yang, Ziyang Wu, Yue Zhao, Yi Ma | Language-Image Alignment with Fixed Text Encoders | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Currently, the most dominant approach to establishing language-image alignment is to pre-train text and image encoders jointly through contrastive learning, such as CLIP and its variants. In this work, we question whether such a costly joint training is necessary. In particular, we investigate if a pre-trained fixed large language model (LLM) offers a good enough text encoder to guide visual representation learning. That is, we propose to learn Language-Image alignment with a Fixed Text encoder (LIFT) from an LLM by training only the image encoder. Somewhat surprisingly, through comprehensive benchmarking and ablation studies, we find that this much simplified framework LIFT is highly effective and it outperforms CLIP in most scenarios that involve compositional understanding and long captions, while achieving considerable gains in computational efficiency. Our work takes a first step towards systematically exploring how text embeddings from LLMs can guide visual learning and suggests an alternative design choice for learning language-aligned visual representations. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:51:56 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yang",
"Jingfeng",
""
],
[
"Wu",
"Ziyang",
""
],
[
"Zhao",
"Yue",
""
],
[
"Ma",
"Yi",
""
]
] |
2506.04210 | Soumya Suvra Ghosal Mr. | Soumya Suvra Ghosal, Souradip Chakraborty, Avinash Reddy, Yifu Lu, Mengdi Wang, Dinesh Manocha, Furong Huang, Mohammad Ghavamzadeh, Amrit Singh Bedi | Does Thinking More always Help? Understanding Test-Time Scaling in Reasoning Models | null | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent trends in test-time scaling for reasoning models (e.g., OpenAI o1, DeepSeek R1) have led to a popular belief that extending thinking traces using prompts like "Wait" or "Let me rethink" can improve performance. This raises a natural question: Does thinking more at test-time truly lead to better reasoning? To answer this question, we perform a detailed empirical study across models and benchmarks, which reveals a consistent pattern of initial performance improvements from additional thinking followed by a decline, due to "overthinking". To understand this non-monotonic trend, we consider a simple probabilistic model, which reveals that additional thinking increases output variance-creating an illusion of improved reasoning while ultimately undermining precision. Thus, observed gains from "more thinking" are not true indicators of improved reasoning, but artifacts stemming from the connection between model uncertainty and evaluation metric. This suggests that test-time scaling through extended thinking is not an effective way to utilize the inference thinking budget. Recognizing these limitations, we introduce an alternative test-time scaling approach, parallel thinking, inspired by Best-of-N sampling. Our method generates multiple independent reasoning paths within the same inference budget and selects the most consistent response via majority vote, achieving up to 20% higher accuracy compared to extended thinking. This provides a simple yet effective mechanism for test-time scaling of reasoning models. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:55:09 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ghosal",
"Soumya Suvra",
""
],
[
"Chakraborty",
"Souradip",
""
],
[
"Reddy",
"Avinash",
""
],
[
"Lu",
"Yifu",
""
],
[
"Wang",
"Mengdi",
""
],
[
"Manocha",
"Dinesh",
""
],
[
"Huang",
"Furong",
""
],
[
"Ghavamzadeh",
"Mohammad",
""
],
[
"Bedi",
"Amrit Singh",
""
]
] |
2506.04211 | Boyong He | Boyong He, Yuxiang Ji, Zhuoyue Tan and Liaoni Wu | Diffusion Domain Teacher: Diffusion Guided Domain Adaptive Object Detector | MM2024 poster, with appendix and codes | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object detectors often suffer a decrease in performance due to the large domain gap between the training data (source domain) and real-world data (target domain). Diffusion-based generative models have shown remarkable abilities in generating high-quality and diverse images, suggesting their potential for extracting valuable feature from various domains. To effectively leverage the cross-domain feature representation of diffusion models, in this paper, we train a detector with frozen-weight diffusion model on the source domain, then employ it as a teacher model to generate pseudo labels on the unlabeled target domain, which are used to guide the supervised learning of the student model on the target domain. We refer to this approach as Diffusion Domain Teacher (DDT). By employing this straightforward yet potent framework, we significantly improve cross-domain object detection performance without compromising the inference speed. Our method achieves an average mAP improvement of 21.2% compared to the baseline on 6 datasets from three common cross-domain detection benchmarks (Cross-Camera, Syn2Real, Real2Artistic}, surpassing the current state-of-the-art (SOTA) methods by an average of 5.7% mAP. Furthermore, extensive experiments demonstrate that our method consistently brings improvements even in more powerful and complex models, highlighting broadly applicable and effective domain adaptation capability of our DDT. The code is available at https://github.com/heboyong/Diffusion-Domain-Teacher. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:56:46 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"He",
"Boyong",
""
],
[
"Ji",
"Yuxiang",
""
],
[
"Tan",
"Zhuoyue",
""
],
[
"Wu",
"Liaoni",
""
]
] |
2506.04212 | Saikat Das | Saikat Das, Siyao Xu, Joonas N\"attil\"a | Studying mirror acceleration via kinetic simulations of relativistic plasma turbulence | 13 pages, 13 figures | null | null | null | astro-ph.HE physics.plasm-ph | http://creativecommons.org/licenses/by/4.0/ | Efficient turbulent acceleration of particles is indicated by recent astrophysical observations, but its mechanism is not well understood. Mirror acceleration has recently been proposed as an efficient mechanism for particle energization in turbulence-compressed magnetic fields. We employ a 3D particle-in-cell (PIC) simulation of pair plasma in magnetized and relativistic turbulence to study this new mechanism and its acceleration efficiency. By tracking individual particles, we see that reversal of a particle's moving direction and significant energy gain can happen during one mirror interaction and within one gyro-orbit. As expected for mirror acceleration, we statistically find that (1) energy gain is preferentially in the direction perpendicular to the local magnetic field and positively correlated with local magnetic field strengthening, and (2) the particle pitch angle distribution becomes increasingly anisotropic toward higher energies, with a concentration at large pitch angles. Our results demonstrate that the mirror acceleration causes a strong confinement of particles by stochastically increasing their pitch angles. This, in turn, facilitates repeated mirror acceleration with the mirroring condition well satisfied. We conclude that mirror acceleration is a promising mechanism accounting for efficient acceleration in magnetized and turbulent astrophysical plasmas. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:56:48 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Das",
"Saikat",
""
],
[
"Xu",
"Siyao",
""
],
[
"Nättilä",
"Joonas",
""
]
] |
2506.04214 | Tingle Li | Tingle Li, Baihe Huang, Xiaobin Zhuang, Dongya Jia, Jiawei Chen, Yuping Wang, Zhuo Chen, Gopala Anumanchipalli, Yuxuan Wang | Sounding that Object: Interactive Object-Aware Image to Audio Generation | ICML 2025 | null | null | null | cs.CV cs.LG cs.MM cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generating accurate sounds for complex audio-visual scenes is challenging, especially in the presence of multiple objects and sound sources. In this paper, we propose an {\em interactive object-aware audio generation} model that grounds sound generation in user-selected visual objects within images. Our method integrates object-centric learning into a conditional latent diffusion model, which learns to associate image regions with their corresponding sounds through multi-modal attention. At test time, our model employs image segmentation to allow users to interactively generate sounds at the {\em object} level. We theoretically validate that our attention mechanism functionally approximates test-time segmentation masks, ensuring the generated audio aligns with selected objects. Quantitative and qualitative evaluations show that our model outperforms baselines, achieving better alignment between objects and their associated sounds. Project page: https://tinglok.netlify.app/files/avobject/ | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:57:26 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Li",
"Tingle",
""
],
[
"Huang",
"Baihe",
""
],
[
"Zhuang",
"Xiaobin",
""
],
[
"Jia",
"Dongya",
""
],
[
"Chen",
"Jiawei",
""
],
[
"Wang",
"Yuping",
""
],
[
"Chen",
"Zhuo",
""
],
[
"Anumanchipalli",
"Gopala",
""
],
[
"Wang",
"Yuxuan",
""
]
] |
2506.04215 | Alex DeWeese | Alex DeWeese, Guannan Qu | Thinking Beyond Visibility: A Near-Optimal Policy Framework for Locally Interdependent Multi-Agent MDPs | null | null | null | null | cs.MA cs.AI cs.LG math.OC | http://creativecommons.org/licenses/by/4.0/ | Decentralized Partially Observable Markov Decision Processes (Dec-POMDPs) are known to be NEXP-Complete and intractable to solve. However, for problems such as cooperative navigation, obstacle avoidance, and formation control, basic assumptions can be made about local visibility and local dependencies. The work DeWeese and Qu 2024 formalized these assumptions in the construction of the Locally Interdependent Multi-Agent MDP. In this setting, it establishes three closed-form policies that are tractable to compute in various situations and are exponentially close to optimal with respect to visibility. However, it is also shown that these solutions can have poor performance when the visibility is small and fixed, often getting stuck during simulations due to the so called "Penalty Jittering" phenomenon. In this work, we establish the Extended Cutoff Policy Class which is, to the best of our knowledge, the first non-trivial class of near optimal closed-form partially observable policies that are exponentially close to optimal with respect to the visibility for any Locally Interdependent Multi-Agent MDP. These policies are able to remember agents beyond their visibilities which allows them to perform significantly better in many small and fixed visibility settings, resolve Penalty Jittering occurrences, and under certain circumstances guarantee fully observable joint optimal behavior despite the partial observability. We also propose a generalized form of the Locally Interdependent Multi-Agent MDP that allows for transition dependence and extended reward dependence, then replicate our theoretical results in this setting. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:57:30 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"DeWeese",
"Alex",
""
],
[
"Qu",
"Guannan",
""
]
] |
2506.04216 | Xuanhua He | Zixuan Ye, Xuanhua He, Quande Liu, Qiulin Wang, Xintao Wang, Pengfei Wan, Di Zhang, Kun Gai, Qifeng Chen, Wenhan Luo | UNIC: Unified In-Context Video Editing | The project page is at \href{https://zixuan-ye.github.io/UNIC}{https://zixuan-ye.github.io/UNIC} | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in text-to-video generation have sparked interest in generative video editing tasks. Previous methods often rely on task-specific architectures (e.g., additional adapter modules) or dedicated customizations (e.g., DDIM inversion), which limit the integration of versatile editing conditions and the unification of various editing tasks. In this paper, we introduce UNified In-Context Video Editing (UNIC), a simple yet effective framework that unifies diverse video editing tasks within a single model in an in-context manner. To achieve this unification, we represent the inputs of various video editing tasks as three types of tokens: the source video tokens, the noisy video latent, and the multi-modal conditioning tokens that vary according to the specific editing task. Based on this formulation, our key insight is to integrate these three types into a single consecutive token sequence and jointly model them using the native attention operations of DiT, thereby eliminating the need for task-specific adapter designs. Nevertheless, direct task unification under this framework is challenging, leading to severe token collisions and task confusion due to the varying video lengths and diverse condition modalities across tasks. To address these, we introduce task-aware RoPE to facilitate consistent temporal positional encoding, and condition bias that enables the model to clearly differentiate different editing tasks. This allows our approach to adaptively perform different video editing tasks by referring the source video and varying condition tokens "in context", and support flexible task composition. To validate our method, we construct a unified video editing benchmark containing six representative video editing tasks. Results demonstrate that our unified approach achieves superior performance on each task and exhibits emergent task composition abilities. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:57:43 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ye",
"Zixuan",
""
],
[
"He",
"Xuanhua",
""
],
[
"Liu",
"Quande",
""
],
[
"Wang",
"Qiulin",
""
],
[
"Wang",
"Xintao",
""
],
[
"Wan",
"Pengfei",
""
],
[
"Zhang",
"Di",
""
],
[
"Gai",
"Kun",
""
],
[
"Chen",
"Qifeng",
""
],
[
"Luo",
"Wenhan",
""
]
] |
2506.04217 | Junting Chen | Junting Chen, Haotian Liang, Lingxiao Du, Weiyun Wang, Mengkang Hu, Yao Mu, Wenhai Wang, Jifeng Dai, Ping Luo, Wenqi Shao, Lin Shao | OWMM-Agent: Open World Mobile Manipulation With Multi-modal Agentic Data Synthesis | 9 pages of main content, 19 pages in total | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | The rapid progress of navigation, manipulation, and vision models has made mobile manipulators capable in many specialized tasks. However, the open-world mobile manipulation (OWMM) task remains a challenge due to the need for generalization to open-ended instructions and environments, as well as the systematic complexity to integrate high-level decision making with low-level robot control based on both global scene understanding and current agent state. To address this complexity, we propose a novel multi-modal agent architecture that maintains multi-view scene frames and agent states for decision-making and controls the robot by function calling. A second challenge is the hallucination from domain shift. To enhance the agent performance, we further introduce an agentic data synthesis pipeline for the OWMM task to adapt the VLM model to our task domain with instruction fine-tuning. We highlight our fine-tuned OWMM-VLM as the first dedicated foundation model for mobile manipulators with global scene understanding, robot state tracking, and multi-modal action generation in a unified model. Through experiments, we demonstrate that our model achieves SOTA performance compared to other foundation models including GPT-4o and strong zero-shot generalization in real world. The project page is at https://github.com/HHYHRHY/OWMM-Agent | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:57:44 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Chen",
"Junting",
""
],
[
"Liang",
"Haotian",
""
],
[
"Du",
"Lingxiao",
""
],
[
"Wang",
"Weiyun",
""
],
[
"Hu",
"Mengkang",
""
],
[
"Mu",
"Yao",
""
],
[
"Wang",
"Wenhai",
""
],
[
"Dai",
"Jifeng",
""
],
[
"Luo",
"Ping",
""
],
[
"Shao",
"Wenqi",
""
],
[
"Shao",
"Lin",
""
]
] |
2506.04218 | Kashyap Chitta | Wei Cao, Marcel Hallgarten, Tianyu Li, Daniel Dauner, Xunjiang Gu, Caojun Wang, Yakov Miron, Marco Aiello, Hongyang Li, Igor Gilitschenski, Boris Ivanovic, Marco Pavone, Andreas Geiger, Kashyap Chitta | Pseudo-Simulation for Autonomous Driving | null | null | null | null | cs.RO cs.AI cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing evaluation paradigms for Autonomous Vehicles (AVs) face critical limitations. Real-world evaluation is often challenging due to safety concerns and a lack of reproducibility, whereas closed-loop simulation can face insufficient realism or high computational costs. Open-loop evaluation, while being efficient and data-driven, relies on metrics that generally overlook compounding errors. In this paper, we propose pseudo-simulation, a novel paradigm that addresses these limitations. Pseudo-simulation operates on real datasets, similar to open-loop evaluation, but augments them with synthetic observations generated prior to evaluation using 3D Gaussian Splatting. Our key idea is to approximate potential future states the AV might encounter by generating a diverse set of observations that vary in position, heading, and speed. Our method then assigns a higher importance to synthetic observations that best match the AV's likely behavior using a novel proximity-based weighting scheme. This enables evaluating error recovery and the mitigation of causal confusion, as in closed-loop benchmarks, without requiring sequential interactive simulation. We show that pseudo-simulation is better correlated with closed-loop simulations (R^2=0.8) than the best existing open-loop approach (R^2=0.7). We also establish a public leaderboard for the community to benchmark new methodologies with pseudo-simulation. Our code is available at https://github.com/autonomousvision/navsim. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:57:53 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Cao",
"Wei",
""
],
[
"Hallgarten",
"Marcel",
""
],
[
"Li",
"Tianyu",
""
],
[
"Dauner",
"Daniel",
""
],
[
"Gu",
"Xunjiang",
""
],
[
"Wang",
"Caojun",
""
],
[
"Miron",
"Yakov",
""
],
[
"Aiello",
"Marco",
""
],
[
"Li",
"Hongyang",
""
],
[
"Gilitschenski",
"Igor",
""
],
[
"Ivanovic",
"Boris",
""
],
[
"Pavone",
"Marco",
""
],
[
"Geiger",
"Andreas",
""
],
[
"Chitta",
"Kashyap",
""
]
] |
2506.04219 | Maria Yampolskaya | Maria Yampolskaya, Laertis Ikonomou, Pankaj Mehta | Finding signatures of low-dimensional geometric landscapes in high-dimensional cell fate transitions | 19 pages, 4 figures | null | null | null | q-bio.CB cond-mat.dis-nn cond-mat.stat-mech | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Multicellular organisms develop a wide variety of highly-specialized cell types. The consistency and robustness of developmental cell fate trajectories suggests that complex gene regulatory networks effectively act as low-dimensional cell fate landscapes. A complementary set of works draws on the theory of dynamical systems to argue that cell fate transitions can be categorized into universal decision-making classes. However, the theory connecting geometric landscapes and decision-making classes to high-dimensional gene expression space is still in its infancy. Here, we introduce a phenomenological model that allows us to identify gene expression signatures of decision-making classes from single-cell RNA-sequencing time-series data. Our model combines low-dimensional gradient-like dynamical systems and high-dimensional Hopfield networks to capture the interplay between cell fate, gene expression, and signaling pathways. We apply our model to the maturation of alveolar cells in mouse lungs to show that the transient appearance of a mixed alveolar type 1/type 2 state suggests the triple cusp decision-making class. We also analyze lineage-tracing data on hematopoetic differentiation and show that bipotent neutrophil-monocyte progenitors likely undergo a heteroclinic flip bifurcation. Our results suggest it is possible to identify universal decision-making classes for cell fate transitions directly from data. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:57:54 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yampolskaya",
"Maria",
""
],
[
"Ikonomou",
"Laertis",
""
],
[
"Mehta",
"Pankaj",
""
]
] |
2506.04220 | Fangrui Zhu | Fangrui Zhu, Hanhui Wang, Yiming Xie, Jing Gu, Tianye Ding, Jianwei Yang, Huaizu Jiang | Struct2D: A Perception-Guided Framework for Spatial Reasoning in Large Multimodal Models | https://github.com/neu-vi/struct2d | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Unlocking spatial reasoning in Large Multimodal Models (LMMs) is crucial for enabling intelligent interaction with 3D environments. While prior efforts often rely on explicit 3D inputs or specialized model architectures, we ask: can LMMs reason about 3D space using only structured 2D representations derived from perception? We introduce Struct2D, a perception-guided prompting framework that combines bird's-eye-view (BEV) images with object marks and object-centric metadata, optionally incorporating egocentric keyframes when needed. Using Struct2D, we conduct an in-depth zero-shot analysis of closed-source LMMs (e.g., GPT-o3) and find that they exhibit surprisingly strong spatial reasoning abilities when provided with structured 2D inputs, effectively handling tasks such as relative direction estimation and route planning. Building on these insights, we construct Struct2D-Set, a large-scale instruction tuning dataset with 200K fine-grained QA pairs across eight spatial reasoning categories, generated automatically from 3D indoor scenes. We fine-tune an open-source LMM (Qwen2.5VL) on Struct2D-Set, achieving competitive performance on multiple benchmarks, including 3D question answering, dense captioning, and object grounding. Our approach demonstrates that structured 2D inputs can effectively bridge perception and language reasoning in LMMs-without requiring explicit 3D representations as input. We will release both our code and dataset to support future research. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:58:04 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Zhu",
"Fangrui",
""
],
[
"Wang",
"Hanhui",
""
],
[
"Xie",
"Yiming",
""
],
[
"Gu",
"Jing",
""
],
[
"Ding",
"Tianye",
""
],
[
"Yang",
"Jianwei",
""
],
[
"Jiang",
"Huaizu",
""
]
] |
2506.04221 | Bowen Shi | Tai-Hsuan Yang, Bowen Shi and Jong Yeon Lee | Topological Mixed States: Axiomatic Approaches and Phases of Matter | 34 pages and 27 figures | null | null | null | cond-mat.str-el cond-mat.stat-mech quant-ph | http://creativecommons.org/licenses/by/4.0/ | For closed quantum systems, topological orders are understood through the equivalence classes of ground states of gapped local Hamiltonians. The generalization of this conceptual paradigm to open quantum systems, however, remains elusive, often relying on operational definitions without fundamental principles. Here, we fill this gap by proposing an approach based on three axioms: ($i$) local recoverability, ($ii$) absence of long-range correlations, and ($iii$) spatial uniformity. States that satisfy these axioms are fixed points; requiring the axioms only after coarse-graining promotes each fixed point to an equivalence class, i.e. a phase, presenting the first step towards the axiomatic classification of mixed-state phases of matter: \emph{mixed-state bootstrap program}.
From these axioms, a rich set of topological data naturally emerges. For example, each topological mixed state supports locally indistinguishable classical and/or quantum logical memories with distinct responses to topological operations. These data label distinct mixed-state phases, allowing one to distinguish them. We further uncover a hierarchy of secret-sharing constraints: in non-Abelian phases, reliable recovery-even of information that looks purely classical -- demands a specific coordination among spatial subregions, a requirement different across non-Abelian classes. This originates from non-Abelian fusion rules that can stay robust under decoherence. Finally, we performed large-scale numerical simulations to corroborate stability-weakly decohered fixed points respect the axioms once coarse-grained. These results lay the foundation for a systematic classification of topological states in open quantum systems. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:58:45 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yang",
"Tai-Hsuan",
""
],
[
"Shi",
"Bowen",
""
],
[
"Lee",
"Jong Yeon",
""
]
] |
2506.04222 | Shikhin Sethi | Shikhin Sethi | Bordered Heegaard Floer modules for satellite operations using planar graphs | 55 pages, 32 figures | null | null | null | math.GT | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Lipshitz, Ozsv\'ath, and Thurston extend the theory of bordered Heegaard Floer homology to compute $\mathbf{CF}^-$. Like with the hat theory, their minus invariants provide a recipe to compute knot invariants associated to satellite knots. We combinatorially construct the weighted $A_\infty$-modules associated to the $(p, 1)$-cable. The operations on these modules count certain classes of inductively constructed decorated planar graphs. This description of the weighted $A_\infty$-modules provides a combinatorial proof of the $A_\infty$ structure relations for the modules. We further prove a uniqueness property for the modules we construct: any weighted extensions of the unweighted $U = 0$ modules have isomorphic associated type D modules. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:58:51 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Sethi",
"Shikhin",
""
]
] |
2506.04223 | Alexis Ralli | Alexis Ralli, Tim Weaving, Peter V. Coveney, Peter J. Love | Bridging Quantum Chemistry and MaxCut: Classical Performance Guarantees and Quantum Algorithms for the Hartree-Fock Method | null | null | null | null | quant-ph cond-mat.mtrl-sci physics.chem-ph physics.comp-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In quantum chemistry, self-consistent field (SCF) algorithms define a nonlinear optimization problem, with both continuous and discrete components. In this work, we derive Hartree-Fock-inspired SCF algorithms that can be exactly written as a sequence of Quadratic Unconstrained Spin/Binary Optimization problems (QUSO/QUBO). We reformulate the optimization problem as a series of MaxCut graph problems, which can be efficiently solved using semi-definite programming techniques. This procedure provides performance guarantees at each SCF step, irrespective of the complexity of the optimization landscape. We numerically demonstrate the QUBO-SCF and MaxCut-SCF methods by studying the hydroxide anion OH- and molecular Nitrogen N2. The largest problem addressed in this study involves a system comprised of 220 qubits (equivalently, spin-orbitals). Our results show that QUBO-SCF and MaxCut-SCF suffer much less from internal instabilities compared with conventional SCF calculations. Additionally, we show that the new SCF algorithms can enhance single-reference methods, such as configuration interaction. Finally, we explore how quantum algorithms for optimization can be applied to the QUSO problems arising from the Hartree-Fock method. Four distinct hybrid-quantum classical approaches are introduced: GAS-SCF, QAOA-SCF, QA-SCF and DQI-SCF. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:02 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ralli",
"Alexis",
""
],
[
"Weaving",
"Tim",
""
],
[
"Coveney",
"Peter V.",
""
],
[
"Love",
"Peter J.",
""
]
] |
2506.04224 | Zirui Wang | Zirui Wang, Wenjing Bian, Xinghui Li, Yifu Tao, Jianeng Wang, Maurice Fallon, Victor Adrian Prisacariu | Seeing in the Dark: Benchmarking Egocentric 3D Vision with the Oxford Day-and-Night Dataset | Project page: https://oxdan.active.vision/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We introduce Oxford Day-and-Night, a large-scale, egocentric dataset for novel view synthesis (NVS) and visual relocalisation under challenging lighting conditions. Existing datasets often lack crucial combinations of features such as ground-truth 3D geometry, wide-ranging lighting variation, and full 6DoF motion. Oxford Day-and-Night addresses these gaps by leveraging Meta ARIA glasses to capture egocentric video and applying multi-session SLAM to estimate camera poses, reconstruct 3D point clouds, and align sequences captured under varying lighting conditions, including both day and night. The dataset spans over 30 $\mathrm{km}$ of recorded trajectories and covers an area of 40,000 $\mathrm{m}^2$, offering a rich foundation for egocentric 3D vision research. It supports two core benchmarks, NVS and relocalisation, providing a unique platform for evaluating models in realistic and diverse environments. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:02 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Wang",
"Zirui",
""
],
[
"Bian",
"Wenjing",
""
],
[
"Li",
"Xinghui",
""
],
[
"Tao",
"Yifu",
""
],
[
"Wang",
"Jianeng",
""
],
[
"Fallon",
"Maurice",
""
],
[
"Prisacariu",
"Victor Adrian",
""
]
] |
2506.04225 | Tianyu Huang | Tianyu Huang, Wangguandong Zheng, Tengfei Wang, Yuhao Liu, Zhenwei Wang, Junta Wu, Jie Jiang, Hui Li, Rynson W.H. Lau, Wangmeng Zuo, Chunchao Guo | Voyager: Long-Range and World-Consistent Video Diffusion for Explorable 3D Scene Generation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Real-world applications like video gaming and virtual reality often demand the ability to model 3D scenes that users can explore along custom camera trajectories. While significant progress has been made in generating 3D objects from text or images, creating long-range, 3D-consistent, explorable 3D scenes remains a complex and challenging problem. In this work, we present Voyager, a novel video diffusion framework that generates world-consistent 3D point-cloud sequences from a single image with user-defined camera path. Unlike existing approaches, Voyager achieves end-to-end scene generation and reconstruction with inherent consistency across frames, eliminating the need for 3D reconstruction pipelines (e.g., structure-from-motion or multi-view stereo). Our method integrates three key components: 1) World-Consistent Video Diffusion: A unified architecture that jointly generates aligned RGB and depth video sequences, conditioned on existing world observation to ensure global coherence 2) Long-Range World Exploration: An efficient world cache with point culling and an auto-regressive inference with smooth video sampling for iterative scene extension with context-aware consistency, and 3) Scalable Data Engine: A video reconstruction pipeline that automates camera pose estimation and metric depth prediction for arbitrary videos, enabling large-scale, diverse training data curation without manual 3D annotations. Collectively, these designs result in a clear improvement over existing methods in visual quality and geometric accuracy, with versatile applications. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:04 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Huang",
"Tianyu",
""
],
[
"Zheng",
"Wangguandong",
""
],
[
"Wang",
"Tengfei",
""
],
[
"Liu",
"Yuhao",
""
],
[
"Wang",
"Zhenwei",
""
],
[
"Wu",
"Junta",
""
],
[
"Jiang",
"Jie",
""
],
[
"Li",
"Hui",
""
],
[
"Lau",
"Rynson W. H.",
""
],
[
"Zuo",
"Wangmeng",
""
],
[
"Guo",
"Chunchao",
""
]
] |
2506.04226 | Akshat Gupta | Akshat Gupta, Maochuan Lu, Thomas Hartvigsen, Gopala Anumanchipalli | Efficient Knowledge Editing via Minimal Precomputation | ACL 2025 Main Conference | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Knowledge editing methods like MEMIT are able to make data and compute efficient updates of factual knowledge by using a single sentence to update facts and their consequences. However, what is often overlooked is a "precomputation step", which requires a one-time but significant computational cost. The authors of MEMIT originally precompute approximately 44 million hidden vectors per edited layer, which requires a forward pass over 44 million tokens. For GPT-J (6B), this precomputation step takes 36 hours on a single GPU, while it takes approximately 40 hours for Llama2-7B. Additionally, this precomputation time grows with model size. In this paper, we show that this excessive computational cost is unnecessary. Knowledge editing using MEMIT and related methods, such as ROME and EMMET, can be performed by pre-computing a very small portion of the 44 million hidden vectors. We first present the theoretical minimum number of hidden vector precomputation required for solutions of these editing methods to exist. We then empirically show that knowledge editing using these methods can be done by pre-computing significantly fewer hidden vectors. Specifically, we show that the precomputation step can be done with less than 0.3% of the originally stipulated number of hidden vectors. This saves a significant amount of precomputation time and allows users to begin editing new models within a few minutes. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:05 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Gupta",
"Akshat",
""
],
[
"Lu",
"Maochuan",
""
],
[
"Hartvigsen",
"Thomas",
""
],
[
"Anumanchipalli",
"Gopala",
""
]
] |
2506.04227 | Zhao-Heng Yin | Zhao-Heng Yin, Sherry Yang, Pieter Abbeel | Object-centric 3D Motion Field for Robot Learning from Human Videos | Project: https://zhaohengyin.github.io/3DMF | null | null | null | cs.RO cs.AI cs.CV cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning robot control policies from human videos is a promising direction for scaling up robot learning. However, how to extract action knowledge (or action representations) from videos for policy learning remains a key challenge. Existing action representations such as video frames, pixelflow, and pointcloud flow have inherent limitations such as modeling complexity or loss of information. In this paper, we propose to use object-centric 3D motion field to represent actions for robot learning from human videos, and present a novel framework for extracting this representation from videos for zero-shot control. We introduce two novel components in its implementation. First, a novel training pipeline for training a ''denoising'' 3D motion field estimator to extract fine object 3D motions from human videos with noisy depth robustly. Second, a dense object-centric 3D motion field prediction architecture that favors both cross-embodiment transfer and policy generalization to background. We evaluate the system in real world setups. Experiments show that our method reduces 3D motion estimation error by over 50% compared to the latest method, achieve 55% average success rate in diverse tasks where prior approaches fail~($\lesssim 10$\%), and can even acquire fine-grained manipulation skills like insertion. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:06 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Yin",
"Zhao-Heng",
""
],
[
"Yang",
"Sherry",
""
],
[
"Abbeel",
"Pieter",
""
]
] |
2506.04228 | Sihui Ji | Sihui Ji, Hao Luo, Xi Chen, Yuanpeng Tu, Yiyang Wang, Hengshuang Zhao | LayerFlow: A Unified Model for Layer-aware Video Generation | Project Page: https://sihuiji.github.io/LayerFlow-Page/ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present LayerFlow, a unified solution for layer-aware video generation. Given per-layer prompts, LayerFlow generates videos for the transparent foreground, clean background, and blended scene. It also supports versatile variants like decomposing a blended video or generating the background for the given foreground and vice versa. Starting from a text-to-video diffusion transformer, we organize the videos for different layers as sub-clips, and leverage layer embeddings to distinguish each clip and the corresponding layer-wise prompts. In this way, we seamlessly support the aforementioned variants in one unified framework. For the lack of high-quality layer-wise training videos, we design a multi-stage training strategy to accommodate static images with high-quality layer annotations. Specifically, we first train the model with low-quality video data. Then, we tune a motion LoRA to make the model compatible with static frames. Afterward, we train the content LoRA on the mixture of image data with high-quality layered images along with copy-pasted video data. During inference, we remove the motion LoRA thus generating smooth videos with desired layers. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:58 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Ji",
"Sihui",
""
],
[
"Luo",
"Hao",
""
],
[
"Chen",
"Xi",
""
],
[
"Tu",
"Yuanpeng",
""
],
[
"Wang",
"Yiyang",
""
],
[
"Zhao",
"Hengshuang",
""
]
] |
2506.04229 | Mor Rozner | Mor Rozner and Enrico Ramirez-Ruiz | Stellar distributions around supermassive black holes in gas-rich nuclear star clusters | null | null | null | null | astro-ph.GA astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the stellar distribution around supermassive black holes (SMBHs) in gas-rich nuclear star clusters (NSCs). NSCs could contain vast amounts of gas, which contribute significantly to shaping the stellar distribution, typically altering the stellar density cusp from the usual Bahcall \& Wolf 1976 solution and consequently affecting the dynamics in the NSC. The dense gaseous environment in NSCs gives rise to dynamical phenomena that are otherwise rare in other gas-free environments. Here we extend the derivation introduced in Bahcall \& Wolf 1976 to include an additional energy dissipation term associated with gas drag. We examine the effect of different forms of gas drag on the stellar density distribution. Finally, we discuss implications on the rates of tidal disruption events and other transients triggered by stellar interactions in gas-rich galactic nuclei. | [
{
"version": "v1",
"created": "Wed, 04 Jun 2025 17:59:59 GMT"
}
] | 2025-06-05T00:00:00 | [
[
"Rozner",
"Mor",
""
],
[
"Ramirez-Ruiz",
"Enrico",
""
]
] |
1610.01993 | Valerio Capraro | Valerio Capraro, Francesca Giardini, Daniele Vilone, Mario Paolucci | Partner selection supported by opaque reputation promotes cooperative behavior | Forthcoming in Judgment and Decision Making | Judgm. decis.
mak. 11 (2016) 589-600 | 10.1017/S1930297500004800 | null | physics.soc-ph q-bio.PE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reputation plays a major role in human societies, and it has been proposed as an explanation for the evolution of cooperation. While the majority of previous studies equates reputation with a transparent and complete history of players' past decisions, in real life, reputations are often ambiguous and opaque. Using web-based experiments, we explore the extent to which opaque reputation works in isolating defectors, with and without partner selection opportunities. Our results show that low reputation works as a signal of untrustworthiness, whereas medium or high reputation are not taken into account by participants for orienting their choices. We also find that reputation without partner selection does not promote cooperative behavior; that is, defectors do not turn into cooperators only for the sake of getting a positive reputation. Finally, in a third study, we find that, when reputation is pivotal to selection, then a substantial proportion of would-be-defectors turn into cooperators. Taken together, these results provide insights on the characteristics of reputation and on the way in which humans make use of it when selecting partners but also when knowing that they will be selected. | [
{
"version": "v1",
"created": "Thu, 06 Oct 2016 18:54:49 GMT"
},
{
"version": "v2",
"created": "Fri, 07 Oct 2016 10:30:47 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Capraro",
"Valerio",
""
],
[
"Giardini",
"Francesca",
""
],
[
"Vilone",
"Daniele",
""
],
[
"Paolucci",
"Mario",
""
]
] |
1705.02351 | Idel Waisberg | GRAVITY Collaboration: I. Waisberg, J. Dexter, O. Pfuhl, R. Abuter, A. Amorin, N. Anugu, J.P. Berger, N. Blind, H. Bonnet, W. Brandner, A. Buron, Y. Cl\'enet, W. de Wit, C. Deen, F. Delplancke-Str\"obele, R. Dembet, G. Duvert, A. Eckart, F. Eisenhauer, P. F\'edou, G. Finger, P. Garcia, R. Garcia Lopez, E. Gendron, R. Genzel, S. Gillessen, X. Haubois, M. Haug, F. Haussmann, Th. Henning, S. Hippler, M. Horrobin, Z. Hubert, L. Jochum, L. Jocou, P. Kervella, Y. Kok, M. Kulas, S. Lacour, V. Lapeyr\`ere, J.-B. Le Bouquin, P. L\`ena, M. Lippa, A. M\'erand, E. M\"uller, T. Ott, L. Pallanca, J. Panduro, T. Paumard, K. Perraut, G. Perrin, S. Rabien, A. Ram\'irez, J. Ramos, C. Rau, R.-R. Rohloff, G. Rousset, J. Sanchez-Bermudez, S. Scheithauer, M. Sch\"oller, C. Straubmeier, E. Sturm, F. Vincent, I. Wank, E. Wieprecht, M. Wiest, E. Wiezorrek, M. Wittkowski, J. Woillez and S. Yazici | Sub-milliarcsecond Optical Interferometry of the HMXB BP Cru with VLTI/GRAVITY | Submitted to ApJ, 22 pages, 17 figures | ApJ 844, 72 (2017) | 10.3847/1538-4357/aa79f1 | null | astro-ph.SR astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We observe the HMXB BP Cru using interferometry in the near-infrared K band with VLTI/GRAVITY. Continuum visibilities are at most partially resolved, consistent with the predicted size of the hypergiant. Differential visibility amplitude ($\Delta |V| \sim 5\%$) and phase ($\Delta \phi \sim 2 \deg$) signatures are observed across the HeI $2.059 \mu$m and Br$\gamma$ lines, the latter seen strongly in emission, unusual for the donor star's spectral type. For a baseline $B \sim 100$m, the differential phase RMS $\sim 0.2 \deg$ corresponds to an astrometric precision of $\sim 2 \mu$as. A model-independent analysis in the marginally resolved limit of interferometry reveals asymmetric and extended emission with a strong wavelength dependence. We propose geometric models based on an extended and distorted wind and/or a high density gas stream, which has long been predicted to be present in this system. The observations show that optical interferometry is now able to resolve HMXBs at the spatial scale at which accretion takes place, and therefore probe the effects of the gravitational and radiation fields of the compact object on its environment. | [
{
"version": "v1",
"created": "Fri, 05 May 2017 18:07:18 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"GRAVITY Collaboration",
"",
""
],
[
"Waisberg",
"I.",
""
],
[
"Dexter",
"J.",
""
],
[
"Pfuhl",
"O.",
""
],
[
"Abuter",
"R.",
""
],
[
"Amorin",
"A.",
""
],
[
"Anugu",
"N.",
""
],
[
"Berger",
"J. P.",
""
],
[
"Blind",
"N.",
""
],
[
"Bonnet",
"H.",
""
],
[
"Brandner",
"W.",
""
],
[
"Buron",
"A.",
""
],
[
"Clénet",
"Y.",
""
],
[
"de Wit",
"W.",
""
],
[
"Deen",
"C.",
""
],
[
"Delplancke-Ströbele",
"F.",
""
],
[
"Dembet",
"R.",
""
],
[
"Duvert",
"G.",
""
],
[
"Eckart",
"A.",
""
],
[
"Eisenhauer",
"F.",
""
],
[
"Fédou",
"P.",
""
],
[
"Finger",
"G.",
""
],
[
"Garcia",
"P.",
""
],
[
"Lopez",
"R. Garcia",
""
],
[
"Gendron",
"E.",
""
],
[
"Genzel",
"R.",
""
],
[
"Gillessen",
"S.",
""
],
[
"Haubois",
"X.",
""
],
[
"Haug",
"M.",
""
],
[
"Haussmann",
"F.",
""
],
[
"Henning",
"Th.",
""
],
[
"Hippler",
"S.",
""
],
[
"Horrobin",
"M.",
""
],
[
"Hubert",
"Z.",
""
],
[
"Jochum",
"L.",
""
],
[
"Jocou",
"L.",
""
],
[
"Kervella",
"P.",
""
],
[
"Kok",
"Y.",
""
],
[
"Kulas",
"M.",
""
],
[
"Lacour",
"S.",
""
],
[
"Lapeyrère",
"V.",
""
],
[
"Bouquin",
"J. -B. Le",
""
],
[
"Lèna",
"P.",
""
],
[
"Lippa",
"M.",
""
],
[
"Mérand",
"A.",
""
],
[
"Müller",
"E.",
""
],
[
"Ott",
"T.",
""
],
[
"Pallanca",
"L.",
""
],
[
"Panduro",
"J.",
""
],
[
"Paumard",
"T.",
""
],
[
"Perraut",
"K.",
""
],
[
"Perrin",
"G.",
""
],
[
"Rabien",
"S.",
""
],
[
"Ramírez",
"A.",
""
],
[
"Ramos",
"J.",
""
],
[
"Rau",
"C.",
""
],
[
"Rohloff",
"R. -R.",
""
],
[
"Rousset",
"G.",
""
],
[
"Sanchez-Bermudez",
"J.",
""
],
[
"Scheithauer",
"S.",
""
],
[
"Schöller",
"M.",
""
],
[
"Straubmeier",
"C.",
""
],
[
"Sturm",
"E.",
""
],
[
"Vincent",
"F.",
""
],
[
"Wank",
"I.",
""
],
[
"Wieprecht",
"E.",
""
],
[
"Wiest",
"M.",
""
],
[
"Wiezorrek",
"E.",
""
],
[
"Wittkowski",
"M.",
""
],
[
"Woillez",
"J.",
""
],
[
"Yazici",
"S.",
""
]
] |
1707.01462 | Ronan Terpereau | J\'er\'emy Blanc, Andrea Fanelli, Ronan Terpereau | Automorphisms of $\mathbb{P}^1$-bundles over rational surfaces | 47 pages, final version | Ãâ°pijournal de Géométrie Algébrique, Volume 6 (January 6, 2023) epiga:7603 | 10.46298/epiga.2023.volume6.7603 | null | math.AG | http://creativecommons.org/licenses/by-sa/4.0/ | In this paper we provide the complete classification of $\mathbb{P}^1$-bundles over smooth projective rational surfaces whose neutral component of the automorphism group is maximal. Our results hold over any algebraically closed field of characteristic zero. | [
{
"version": "v1",
"created": "Wed, 05 Jul 2017 16:50:15 GMT"
},
{
"version": "v2",
"created": "Tue, 19 Sep 2017 16:12:40 GMT"
},
{
"version": "v3",
"created": "Sun, 29 Aug 2021 08:53:03 GMT"
},
{
"version": "v4",
"created": "Mon, 27 Jun 2022 17:21:10 GMT"
},
{
"version": "v5",
"created": "Mon, 02 Jan 2023 13:11:47 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Blanc",
"Jérémy",
""
],
[
"Fanelli",
"Andrea",
""
],
[
"Terpereau",
"Ronan",
""
]
] |
1711.07409 | Mirko R\"osner | Mirko R\"osner, Rainer Weissauer | Regular poles for spinor $L$-series attached to split Bessel models of $\mathrm{GSp}(4)$ | 83 pages, 7 tables | Pacific J. Math. 337 (2025) 137-199 | 10.2140/pjm.2025.337.137 | null | math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | For irreducible smooth representations $\Pi$ of $\mathrm{GSp}(4,k)$ over a non-archimedean local field $k$, Piatetskii-Shapiro and Soudry have constructed an $L$-factor depending on the choice of a Bessel model. It factorizes into a regular part and an exceptional part. We determine the regular part for the case of split Bessel models. | [
{
"version": "v1",
"created": "Mon, 20 Nov 2017 16:58:01 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Sep 2018 16:33:54 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Rösner",
"Mirko",
""
],
[
"Weissauer",
"Rainer",
""
]
] |
1801.05510 | Igor V. Nikolaev | Andrey Yu. Glubokov and Igor V. Nikolaev | Jones Index Theorem revisited | to appear in Journal of Pure and Applied Algebra | null | null | null | math.OA math.QA math.RT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We prove the Jones Index Theorem using the K-theory of a cluster $C^*$-algebra of the Riemann sphere with two boundary components. | [
{
"version": "v1",
"created": "Wed, 17 Jan 2018 00:19:54 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Jan 2022 00:05:41 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 11:02:34 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Glubokov",
"Andrey Yu.",
""
],
[
"Nikolaev",
"Igor V.",
""
]
] |
1802.08198 | Idel Waisberg | Idel Waisberg, Jason Dexter, Stefan Gillessen, Oliver Pfuhl, Frank Eisenhauer, Phillip M. Plewa, Michi Baub\"ock, Alejandra Jimenez-Rosales, Maryam Habibi, Thomas Ott, Sebastiano von Fellenberg, Feng Gao, Felix Widmann and Reinhard Genzel | What stellar orbit is needed to measure the spin of the Galactic center black hole from astrometric data? | Accepted to MNRAS | MNRAS 476, 3600 (2018) | 10.1093/mnras/sty476 | null | astro-ph.GA gr-qc | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Astrometric and spectroscopic monitoring of individual stars orbiting the supermassive black hole in the Galactic Center offer a promising way to detect general relativistic effects. While low-order effects are expected to be detected following the periastron passage of S2 in Spring 2018, detecting higher-order effects due to black hole spin will require the discovery of closer stars. In this paper, we set out to determine the requirements such a star would have to satisfy to allow the detection of black hole spin. We focus on the instrument GRAVITY, which saw first light in 2016 and which is expected to achieve astrometric accuracies $10-100 \mu$as. For an observing campaign with duration $T$ years, $N_{obs}$ total observations, astrometric precision $\sigma_x$ and normalized black hole spin $\chi$, we find that $a_{orb}(1-e^2)^{3/4} \lesssim 300 R_S \sqrt{\frac{T}{4 \text{years}}} \left(\frac{N_{obs}}{120}\right)^{0.25} \sqrt{\frac{10 \mu as}{\sigma_x}} \sqrt{\frac{\chi}{0.9}}$ is needed. For $\chi=0.9$ and a potential observing campaign with $\sigma_x = 10 \mu$as, 30 observations/year and duration 4-10 years, we expect $\sim 0.1$ star with $K<19$ satisfying this constraint based on the current knowledge about the stellar population in the central 1". We also propose a method through which GRAVITY could potentially measure radial velocities with precision $\sim 50$ km/s. If the astrometric precision can be maintained, adding radial velocity information increases the expected number of stars by roughly a factor of two. While we focus on GRAVITY, the results can also be scaled to parameters relevant for future extremely large telescopes. | [
{
"version": "v1",
"created": "Thu, 22 Feb 2018 17:44:24 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Waisberg",
"Idel",
""
],
[
"Dexter",
"Jason",
""
],
[
"Gillessen",
"Stefan",
""
],
[
"Pfuhl",
"Oliver",
""
],
[
"Eisenhauer",
"Frank",
""
],
[
"Plewa",
"Phillip M.",
""
],
[
"Bauböck",
"Michi",
""
],
[
"Jimenez-Rosales",
"Alejandra",
""
],
[
"Habibi",
"Maryam",
""
],
[
"Ott",
"Thomas",
""
],
[
"von Fellenberg",
"Sebastiano",
""
],
[
"Gao",
"Feng",
""
],
[
"Widmann",
"Felix",
""
],
[
"Genzel",
"Reinhard",
""
]
] |
1811.00204 | Kiran S. Kedlaya | Kiran S. Kedlaya | Etale and crystalline companions, I | 30 pages; v6: published version | Ãâ°pijournal de Géométrie Algébrique, Volume 6 (December 2, 2022) epiga:6820 | 10.46298/epiga.2022.6820 | null | math.NT math.AG | http://creativecommons.org/licenses/by-sa/4.0/ | Let $X$ be a smooth scheme over a finite field of characteristic $p$. Consider the coefficient objects of locally constant rank on $X$ in $\ell$-adic Weil cohomology: these are lisse Weil sheaves in \'etale cohomology when $\ell \neq p$, and overconvergent $F$-isocrystals in rigid cohomology when $\ell=p$. Using the Langlands correspondence for global function fields in both the \'etale and crystalline settings (work of Lafforgue and Abe, respectively), one sees that on a curve, any coefficient object in one category has "companions" in the other categories with matching characteristic polynomials of Frobenius at closed points. A similar statement is expected for general $X$; building on work of Deligne, Drinfeld showed that any \'etale coefficient object has \'etale companions. We adapt Drinfeld's method to show that any crystalline coefficient object has \'etale companions; this has been shown independently by Abe--Esnault. We also prove some auxiliary results relevant for the construction of crystalline companions of \'etale coefficient objects; this subject will be pursued in a subsequent paper. | [
{
"version": "v1",
"created": "Thu, 01 Nov 2018 03:21:32 GMT"
},
{
"version": "v2",
"created": "Tue, 07 May 2019 18:39:43 GMT"
},
{
"version": "v3",
"created": "Mon, 24 Aug 2020 03:38:41 GMT"
},
{
"version": "v4",
"created": "Fri, 27 Aug 2021 21:02:20 GMT"
},
{
"version": "v5",
"created": "Thu, 28 Jul 2022 15:31:49 GMT"
},
{
"version": "v6",
"created": "Thu, 01 Dec 2022 16:03:00 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Kedlaya",
"Kiran S.",
""
]
] |
1811.12558 | Idel Waisberg | Idel Waisberg, Jason Dexter, Pierre Olivier-Petrucci, Guillaume Dubus, and Karine Perraut | Super-Keplerian Equatorial Outflows in SS 433. Centrifugal Ejection of the Circumbinary Disk | submitted to A&A | A&A 623, A47 (2019) | 10.1051/0004-6361/201834746 | null | astro-ph.SR astro-ph.GA astro-ph.HE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The microquasar SS 433 is the only known steady supercritical accretor in the Galaxy. It is well-known for its relativistic baryonic jets, but the system also drives equatorial outflows. These have been routinely detected in radio images, and components associated with a circumbinary disk have also been suggested in optical emission lines. We wish to spatially resolve the regions producing the stationary emission lines of SS 433 to shed light on its circumbinary structure and outflows. With an estimated binary orbit size $\lesssim 0.1 \text{ mas}$, this requires optical interferometry. We use the optical interferometer VLTI+GRAVITY to spatially resolve SS 433 in the near-infrared K band at high spectral resolution ($R\approx 4000$) on three nights in July 2017. The stationary Br$\gamma$ line is clearly dominated by an extended $\sim 1 \text{ mas} \sim 5 \text{ AU}$ circumbinary structure perpendicular to the jets and with a strong rotation component. The rotation direction is retrograde relative to the jet precession, in accordance with the slaved disk precession model. The structure has a very high specific angular momentum and is too extended to be a stable circumbinary disk in Keplerian rotation; interpreting it as such leads to a very high enclosed mass $M \gtrsim 400 M_{\odot}$. We instead interpret it as the centrifugal ejection of the circumbinary disk, with the implication that there must be an efficient transfer of specific angular momentum from the binary to the disk. We suggest that the equatorial outflows sometimes seen in radio images result from similar episodes of circumbinary disk centrifugal ejection. In addition to the equatorial structure, we find a very extended $\sim 6 \text{ mas} \sim 30 \text{ AU}$ spherical wind component to the Br$\gamma$ line: the entire binary is engulfed in an optically thin spherical line emission envelope. | [
{
"version": "v1",
"created": "Fri, 30 Nov 2018 00:45:06 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Waisberg",
"Idel",
""
],
[
"Dexter",
"Jason",
""
],
[
"Olivier-Petrucci",
"Pierre",
""
],
[
"Dubus",
"Guillaume",
""
],
[
"Perraut",
"Karine",
""
]
] |
1811.12564 | Idel Waisberg | Idel Waisberg, Jason Dexter, Pierre Olivier-Petrucci, Guillaume Dubus, and Karine Perraut | The Collimated Radiation in SS 433. Constraints from Spatially Resolved Optical Jets and $\texttt{Cloudy}$ Modeling of the Optical Bullets | submitted to A&A | A&A 624, A127 (2019) | 10.1051/0004-6361/201834747 | null | astro-ph.HE astro-ph.GA astro-ph.SR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The microquasar SS 433 is well-known for its precessing, relativistic baryonic jets. Depending on their heating mechanism, the optical jet bullets may serve as a probe of the collimated radiation coming from the inner region close to the compact object. The optical interferometer VLTI/GRAVITY has allowed to spatially resolved the optical jets in SS 433 for the first time. We present here the second such observation taken over three nights in July 2017. In addition, we use the multi-wavelength XSHOOTER spectrograph at VLT to study the optical bullets in SS 433 in detail. GRAVITY reveals elongated exponential-like spatial profiles for the optical jets, suggestive of a heating mechanism acting throughout a long portion of the jet and naturally explained by photoionization by the collimated radiation. We also spatially resolve the movement of the optical bullets for the first time, detecting extended jet components corresponding to previous ejections. \texttt{Cloudy} photoionization models explain both the spatial intensity profiles measured with GRAVITY and the line ratios from XSHOOTER, and constrain the properties of the optical bullets and the ionizing radiation. We find that the latter must peak in the UV with an isotropic luminosity (as inferred by a face on observer) $\approx 10^{41}$ erg/s. Provided that the X-ray SED is sufficiently hard, the collimated X-ray luminosity could still be high enough so that the face on observer would see SS 433 as an ULX ($L_X \lesssim 10^{40}$ erg/s) and it would still be compatible with the H/He/He+ ionization balance of the optical bullets. The kinetic power in the optical jets is constrained to $3-20 \times 10^{38}$ erg/s, and the extinction in the optical jets to $A_V = 6.7 \pm 0.1$. We suggest there may be substantial $A_V \gtrsim 1$ and structured circumstellar extinction in SS 433, likely arising from dust formed in equatorial outflows. | [
{
"version": "v1",
"created": "Fri, 30 Nov 2018 01:05:50 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Waisberg",
"Idel",
""
],
[
"Dexter",
"Jason",
""
],
[
"Olivier-Petrucci",
"Pierre",
""
],
[
"Dubus",
"Guillaume",
""
],
[
"Perraut",
"Karine",
""
]
] |
1812.00830 | Mohsen Asgharzadeh | Mohsen Asgharzadeh | Reflexivity revisited | null | null | null | null | math.AC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study some aspects of reflexive modules. For example, we search conditions for which reflexive modules are free or being very close to free modules. | [
{
"version": "v1",
"created": "Mon, 03 Dec 2018 15:22:43 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Dec 2018 15:27:32 GMT"
},
{
"version": "v3",
"created": "Wed, 30 Jan 2019 13:19:04 GMT"
},
{
"version": "v4",
"created": "Mon, 23 Dec 2019 14:49:48 GMT"
},
{
"version": "v5",
"created": "Mon, 30 May 2022 14:35:28 GMT"
},
{
"version": "v6",
"created": "Wed, 14 Dec 2022 14:29:16 GMT"
},
{
"version": "v7",
"created": "Tue, 03 Jun 2025 13:58:00 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Asgharzadeh",
"Mohsen",
""
]
] |
1901.02314 | Valerio Capraro | Valerio Capraro and Andrea Vanzo | The power of moral words: Loaded language generates framing effects in the extreme dictator game | Forthcoming in Judgment and Decision Making | Judgm. decis.
mak. 14 (2019) 309-317 | 10.1017/S1930297500004356 | null | physics.soc-ph cs.GT q-bio.PE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding whether preferences are sensitive to the frame has been a major topic of debate in the last decades. For example, several works have explored whether the dictator game in the give frame gives rise to a different rate of pro-sociality than the same game in the take frame, leading to mixed results. Here we contribute to this debate with two experiments. In Study 1 ($N=567$) we implement an extreme dictator game in which the dictator either gets \$0.50 and the recipient gets nothing, or the opposite (i.e., the recipient gets \$0.50 and the dictator gets nothing). We experimentally manipulate the words describing the available actions using six terms, from very negative (e.g., stealing) to very positive (e.g., donating) connotations. We find that the rate of pro-sociality is affected by the words used to describe the available actions. In Study 2 ($N=221$) we ask brand new participants to rate each of the words used in Study 1 from ``extremely wrong'' to ``extremely right'' . We find that these moral judgments explain the framing effect in Study 1. In sum, our studies provide evidence that framing effects in an extreme Dictator game can be generated using morally loaded language. | [
{
"version": "v1",
"created": "Tue, 08 Jan 2019 14:08:36 GMT"
},
{
"version": "v2",
"created": "Wed, 13 Feb 2019 11:36:42 GMT"
},
{
"version": "v3",
"created": "Sat, 06 Apr 2019 16:19:49 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Capraro",
"Valerio",
""
],
[
"Vanzo",
"Andrea",
""
]
] |
1903.08885 | Simone Marchesi | Simone Marchesi (UB), Jean Vall\`es (UPPA) | Triangular arrangements on the projective plane | 20 pages. Published in \'Epijournal de G\'eom\'etrie Alg\'ebrique. Section 4 has been deeply revised due to an incorrect statement pointed out by the referee | Ãâ°pijournal de Géométrie Algébrique, Volume 7 (May 2, 2023) epiga:7323 | 10.46298/epiga.2023.7323 | null | math.AG math.AT math.CO | http://creativecommons.org/licenses/by-sa/4.0/ | In this work we study line arrangements consisting in lines passing through three non-aligned points. We call them triangular arrangements. We prove that any combinatorics of a triangular arrangement is always realized by a Roots-of-Unity-Arrangement, which is a particular class of triangular arrangements. Among these Roots-of Unity-Arrangements, we provide conditions that ensure their freeness. Finally, we give two triangular arrangements having the same weak combinatorics, such that one is free but the other one is not. | [
{
"version": "v1",
"created": "Thu, 21 Mar 2019 09:06:54 GMT"
},
{
"version": "v2",
"created": "Fri, 12 Apr 2019 09:22:35 GMT"
},
{
"version": "v3",
"created": "Tue, 16 Apr 2019 00:07:53 GMT"
},
{
"version": "v4",
"created": "Sat, 05 Oct 2019 15:48:08 GMT"
},
{
"version": "v5",
"created": "Wed, 31 Mar 2021 17:49:37 GMT"
},
{
"version": "v6",
"created": "Sun, 30 Apr 2023 10:05:15 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Marchesi",
"Simone",
"",
"UB"
],
[
"Vallès",
"Jean",
"",
"UPPA"
]
] |
1907.12578 | Antony Maciocia | Marcos Jardim and Antony Maciocia | Walls and asymptotics for Bridgeland stability conditions on 3-folds | 61 pages with 11 figures. Published in \'Epijournal de G\'eom\'etrie Alg\'ebrique | Ãâ°pijournal de Géométrie Algébrique, Volume 6 (December 22, 2022) epiga:6819 | 10.46298/epiga.2022.6819 | null | math.AG math.CT math.DG | http://creativecommons.org/licenses/by-sa/4.0/ | We consider Bridgeland stability conditions for three-folds conjectured by Bayer-Macr\`i-Toda in the case of Picard rank one. We study the differential geometry of numerical walls, characterizing when they are bounded, discussing possible intersections, and showing that they are essentially regular. Next, we prove that walls within a certain region of the upper half plane that parametrizes geometric stability conditions must always intersect the curve given by the vanishing of the slope function and, for a fixed value of s, have a maximum turning point there. We then use all of these facts to prove that Gieseker semistability is equivalent to asymptotic semistability along a class of paths in the upper half plane, and to show how to find large families of walls. We illustrate how to compute all of the walls and describe the Bridgeland moduli spaces for the Chern character (2,0,-1,0) on complex projective 3-space in a suitable region of the upper half plane. | [
{
"version": "v1",
"created": "Mon, 29 Jul 2019 18:02:16 GMT"
},
{
"version": "v2",
"created": "Thu, 01 Oct 2020 11:46:08 GMT"
},
{
"version": "v3",
"created": "Wed, 18 May 2022 16:05:18 GMT"
},
{
"version": "v4",
"created": "Tue, 20 Dec 2022 19:49:14 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Jardim",
"Marcos",
""
],
[
"Maciocia",
"Antony",
""
]
] |
1908.09164 | Gabriel Angelini-Knoll | Gabriel Angelini-Knoll and J.D. Quigley | A homological approach to chromatic complexity of algebraic K-theory | 36 pages. Significantly revised after referee feedback. Accepted for publication at JPAA | null | null | null | math.AT math.KT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The family of Thom spectra $y(n)$ interpolates between the sphere spectrum and the mod two Eilenberg--MacLane spectrum. Computations of Mahowald, Ravenel, Shick, and the authors show that the associative ring spectrum $y(n)$ has type $n$. Using trace methods, we give evidence that algebraic K-theory preserves this chromatic complexity. Our approach sheds light on the chromatic complexity of topological negative cyclic homology and topological periodic cyclic homology, which approximate algebraic K-theory and are of independent interest. Our main contribution is a homological approach that can be applied in great generality, such as to associative ring spectra $R$ without additional structure whose coefficient rings are not completely understood. | [
{
"version": "v1",
"created": "Sat, 24 Aug 2019 16:34:10 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Mar 2020 01:48:51 GMT"
},
{
"version": "v3",
"created": "Tue, 03 Jun 2025 14:15:36 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Angelini-Knoll",
"Gabriel",
""
],
[
"Quigley",
"J. D.",
""
]
] |
1908.10986 | Franco Rota | Matteo Altavilla, Marin Petkovic, Franco Rota | Moduli spaces on the Kuznetsov component of Fano threefolds of index 2 | V3: 34 pages, substantially rewritten to improve exposition, references updated. Comments welcome! V4: 31 pages, final version | Ãâ°pijournal de Géométrie Algébrique, Volume 6 (July 8, 2022) epiga:7047 | 10.46298/epiga.2022.7047 | null | math.AG | http://creativecommons.org/licenses/by/4.0/ | General hyperplane sections of a Fano threefold $Y$ of index 2 and Picard rank 1 are del Pezzo surfaces, and their Picard group is related to a root system. To the corresponding roots, we associate objects in the Kuznetsov component of $Y$ and investigate their moduli spaces, using the stability condition constructed by Bayer, Lahoz, Macr\`i, and Stellari, and the Abel--Jacobi map. We identify a subvariety of the moduli space isomorphic to $Y$ itself, and as an application we prove a (refined) categorical Torelli theorem for general quartic double solids. | [
{
"version": "v1",
"created": "Wed, 28 Aug 2019 23:30:20 GMT"
},
{
"version": "v2",
"created": "Wed, 04 Sep 2019 18:22:21 GMT"
},
{
"version": "v3",
"created": "Fri, 01 Jan 2021 17:26:44 GMT"
},
{
"version": "v4",
"created": "Wed, 06 Jul 2022 21:23:43 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Altavilla",
"Matteo",
""
],
[
"Petkovic",
"Marin",
""
],
[
"Rota",
"Franco",
""
]
] |
2001.04426 | Ya Deng | Ya Deng | Big Picard theorems and algebraic hyperbolicity for varieties admitting a variation of Hodge structures | 31 pages. Final version, to appear in \'Epijournal de G\'eom\'etrie Alg\'ebrique | Ãâ°pijournal de Géométrie Algébrique, Volume 7 (April 24, 2023) epiga:8393 | 10.46298/epiga.2023.volume7.8393 | null | math.AG math.CV | http://creativecommons.org/licenses/by-sa/4.0/ | In this paper, we study various hyperbolicity properties for a quasi-compact K\"ahler manifold $U$ which admits a complex polarized variation of Hodge structures so that each fiber of the period map is zero-dimensional. In the first part, we prove that $U$ is algebraically hyperbolic and that the generalized big Picard theorem holds for $U$. In the second part, we prove that there is a finite \'etale cover $\tilde{U}$ of $U$ from a quasi-projective manifold $\tilde{U}$ such that any projective compactification $X$ of $\tilde{U}$ is Picard hyperbolic modulo the boundary $X-\tilde{U}$, and any irreducible subvariety of $X$ not contained in $X-\tilde{U}$ is of general type. This result coarsely incorporates previous works by Nadel, Rousseau, Brunebarbe and Cadorel on the hyperbolicity of compactifications of quotients of bounded symmetric domains by torsion-free lattices. | [
{
"version": "v1",
"created": "Mon, 13 Jan 2020 17:52:00 GMT"
},
{
"version": "v2",
"created": "Tue, 04 Feb 2020 15:28:43 GMT"
},
{
"version": "v3",
"created": "Thu, 16 Apr 2020 14:48:24 GMT"
},
{
"version": "v4",
"created": "Sun, 13 Jun 2021 09:24:31 GMT"
},
{
"version": "v5",
"created": "Thu, 20 Apr 2023 19:57:56 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Deng",
"Ya",
""
]
] |
2001.10654 | Hsueh-Yung Lin | Hsueh-Yung Lin | On the dual positive cones and the algebraicity of a compact K\"ahler manifold | Published version | Ãâ°pijournal de Géométrie Algébrique, Special volume in honour of Claire Voisin (February 2, 2024) epiga:10771 | 10.46298/epiga.2024.10771 | null | math.AG math.CV | http://creativecommons.org/licenses/by-sa/4.0/ | We investigate the algebraicity of compact K\"ahler manifolds admitting a positive rational Hodge class of bidimension $(1,1)$. We prove that if the dual K\"ahler cone of a compact K\"ahler manifold $X$ contains a rational class as an interior point, then its Albanese variety is projective. As a consequence, we answer the Oguiso--Peternell problem for Ricci-flat compact K\"ahler manifolds. We also study related algebraicity problems for threefolds. | [
{
"version": "v1",
"created": "Wed, 29 Jan 2020 01:08:28 GMT"
},
{
"version": "v2",
"created": "Mon, 06 Sep 2021 15:02:28 GMT"
},
{
"version": "v3",
"created": "Wed, 04 Jan 2023 11:43:03 GMT"
},
{
"version": "v4",
"created": "Tue, 30 Jan 2024 03:04:24 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Lin",
"Hsueh-Yung",
""
]
] |
2002.01371 | Adan Cabello | Luciano Pereira, Alejandro Rojas, Gustavo Lima, Aldo Delgado, Gustavo Ca\~nas, and Ad\'an Cabello | Minimum optical depth multiport interferometers for approximating arbitrary unitary operations and pure states | 8 pages, 4 figures | Phys. Rev. A 111, 062603 (2025) | 10.1103/PhysRevA.111.062603 | null | quant-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reconfigurable devices which can implement arbitrary unitary operations are crucial for photonic quantum computation, optical neural networks, and boson sampling. Here, we address the problem of, using multiport interferometers, approximating with a given infidelity any pure state preparation and any unitary operation. By means of numerical calculations, we show that pure states, in any dimension $d$, can be prepared with infidelity $\le 10^{-15}$ with three layers of $d$ dimensional Fourier transforms and three layers of configurable phase shifters. We also present numerical evidence that $d+1$ layers of $d$-dimensional Fourier transforms and $d+2$ layers of configurable phase shifters can produce any unitary operation with infidelity $\le 10^{-14}$. The conclusions are achieved by numerical simulations in the range from $d = 3$ to $d=10$. These results are interesting in light of the recent availability of multicore fiber-integrated multiport interferometers. | [
{
"version": "v1",
"created": "Tue, 04 Feb 2020 15:40:49 GMT"
},
{
"version": "v2",
"created": "Thu, 05 Aug 2021 13:52:54 GMT"
},
{
"version": "v3",
"created": "Fri, 06 Aug 2021 16:11:56 GMT"
},
{
"version": "v4",
"created": "Tue, 10 May 2022 19:20:44 GMT"
},
{
"version": "v5",
"created": "Mon, 02 Jun 2025 21:59:28 GMT"
}
] | 2025-06-04T00:00:00 | [
[
"Pereira",
"Luciano",
""
],
[
"Rojas",
"Alejandro",
""
],
[
"Lima",
"Gustavo",
""
],
[
"Delgado",
"Aldo",
""
],
[
"Cañas",
"Gustavo",
""
],
[
"Cabello",
"Adán",
""
]
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.