Search arXiv⌕ Search

arXiv subjects

Search papers

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

At least 577 records · Page 32Linked to original sources
arXiv2026

Frobenius functors and $n$-torsionfree objects

We study $n$-torsionfree objects in abelian categories with enough projectives. Frobenius functors preserve $n$-torsionfreeness, and faithful ones reflect it. We prove that stabilization of the torsionfree filtration implies weak Gorensteinness. For Frobenius extensions satisfying a generator condition, we compare the terms of minimal injective resolutions and obtain transfer of Auslander-type conditions and of the Auslander--Gorenstein conjecture. We also compute a family of non-Gorenstein algebras whose torsionfree filtrations stabilize at level two and contain explicit nonprojective Gorenstein projective modules.

math.RT↗
arXiv2026

Suppression of Blow-up in Two-dimensional Keller--Segel Systems by Stochastic Couette Flows

We study Keller--Segel systems on $ \mathbb{T}\times\mathbb{R}$ subject to a stochastic Couette transport flow. We prove that sufficiently strong mixing suppresses chemotactic finite-time blow-up with high probability; thus yielding a unique global-in-time mild solution for initial data of arbitrary mass. Moreover, we obtain a quantitative estimate for the probability of blow-up, showing that it decays exponentially with increasing strength of the stochastic shear. Our analysis is based on the enhanced dissipation generated by the stochastic shear flow. To prove our main result, we first construct a maximal local mild solution through pathwise estimates for the random solution operator of passive scalar transport equations and a fixed-point argument. To obtain global control, we decompose the evolution into time blocks and exploit mixing-induced decay of the nonzero spatial Fourier modes. A nonlinear bootstrap argument then yields global existence above an almost-sure finite random threshold for the shear strength which is obtained by an application of the Borel-Cantelli theorem.

math.AP↗
arXiv2026

IndustrialVLA-Bench: A Traceable Multi-Axis Evaluation of Open Robot Policy Models

Open robot policies increasingly follow two paradigms: vision-language-action models (VLAs) directly map observations and instructions to actions, whereas world-action models (WAMs) incorporate learned video or world dynamics into policy learning or action generation. Although both target the same manipulation tasks and represent alternative design choices, they are commonly reported under different evaluation protocols, leaving their capability, robustness, language sensitivity, and deployment-cost trade-offs unclear. We present IndustrialVLA-Bench, an evidence-aware evaluation of six released VLA and WAM systems under a unified reporting schema. It separately evaluates clean capability on LIBERO, non-language robustness on LIBERO-Plus, instruction sensitivity on LIBERO-Para, and observed execution cost. Reported task scores aggregate three complete evaluations with distinct random seeds under a fixed checkpoint and inference configuration. Across all six systems, clean LIBERO averages differ by only 1.58 points, whereas robustness and paraphrase summaries span 14.62 and 31.08 points. Restricting every comparison to the three protocol-faithful systems preserves the effect (1.36, 14.62 and 23.10 points), so the diagnostic separation reported here does not depend on the weaker evidence tiers. We additionally report observed inference latency, peak memory, runtime mode, and an evidence status for every system. Protocol-faithful, near-reproduction, and pending-verification entries remain visibly separated; only protocol-faithful entries support strict comparisons. Rather than claiming universal superiority of either paradigm, IndustrialVLA-Bench provides traceable evidence for comparing released robot policies on shared practical criteria. Code and evaluation records are available at https://github.com/xiaoqi-7/IndustrialVLA-Bench.

cs.RO↗
arXiv2026

Generalists Act, Specialists Intervene: Modular Stage-Selective Reinforcement Learning for Vision-Language-Action Manipulation

Vision-language-action (VLA) models often struggle in the precision-critical phases of multi-stage manipulation tasks. To mitigate this issue, VLA models can be used in conjunction with reinforcement learning (RL) specialists that are specifically trained to handle the precision-critical phases. However, the coordination between the base VLA model and the RL specialists, which dictates when a specialist should take over from the base VLA and vice-versa, remains an open research question. In this paper, we address this gap by introducing RouteRLT, a modular framework that coordinates a generalist VLA, used as the default controller, with designated precision-critical RL specialists. At a high level, our framework trains a phase-aware coordination mechanism that handles handoffs between the generalist and the specialists. We evaluate RouteRLT on the LIBERO and LIBERO-Plus benchmarks, as well as on a physical connector pickup and insertion task. Overall, we find that RouteRLT improves success on LIBERO, and retains net gains on LIBERO-Plus. On the physical task, RouteRLT completes 65.7% of trials, compared with 8.6% for the baseline. Altogether, these results demonstrate that learned coordination builds on generalist VLA capabilities to improve task completion in precision-critical manipulation.

cs.RO↗
arXiv2026

Gravitational-Susceptibility Dip Behind DESI's results: Addressing the Phantom Crossing by Local limit of Nonlocal Gravity

The second data release of the Dark Energy Spectroscopic Instrument (DESI), based on more than fourteen million galaxies and quasars, prefers a dark energy component whose equation of state is dynamical. In the simplest two-parameter description of this behavior, the Chevallier--Polarski--Linder (CPL) parametrization, the data favor the quadrant $w_0>-1$, $w_a<0$, which implies that the equation of state crosses the phantom divide $w=-1$ at a redshift $z\simeq0.4$. We show that the local limit of nonlocal gravity, a teleparallel extension of general relativity governed by a scalar function called gravitational susceptibility $S(x)$, offers a natural explanation. In this model a cosmological constant is not allowed; dark energy is necessarily dynamical. The phantom crossing seen by an observer is an effective phenomenon generated by the time evolution of $S(z)$. We establish a correspondence: to every general-relativistic dark energy model with equation of state $w(z)$, there corresponds a modified TEGR cosmology with $w=-1$ and a suitable susceptibility $S(z)$ that are indistinguishable at the level of the background expansion. We construct the susceptibility functions that reproduce the CPL behavior favored by DESI data, and we confront the model with the combined CMB,BAO, and SNe data using Boltzmann solver codes. The recovered susceptibility displays a characteristic low-redshift dip with depth $β= -0.025 \pm 0.007$ at $z\simeq 0.4 $, and its derivative changes sign at a redshift equal to the phantom-crossing epoch.

astro-ph.CO↗
arXiv2026

Type-Safe Is Not Error-Free: Typed Decision Models Follow the Option Name, Not the Definition Bound to It

Typed decision models return structured results, but output-type correctness alone does not ensure that decisions follow explicit option definitions. Each option pairs a name with a definition that defines its intended meaning; the name, however, can provide a competing semantic cue. We study this conflict in Jev and two open-weight models by changing only the name-definition mapping, leaving the question, state, and the names and definition texts themselves unchanged. We measure decision flips at the level of the selected definition, rather than the returned name. On 1200 decision tasks with task-specific definitions, decision-flip rates are up to 70.4 pp higher with yes/no names than with the 0/1 control. This gap holds across all 4 binary decision rules. With yes/no names, reassignment also lowers their mean AUC from 93.8% to a below-chance 23.2%. In the binary evaluations, random strings used as option names yield mean flip rates close to those of neutral controls across all three models, with comparable balanced accuracy before reassignment. Together, these results support option-name polarity as a contributor to decision instability beyond reassignment alone. The type-error rate remains 0% throughout, showing that type-correct outputs can still fail to follow explicit option definitions.

cs.AI↗
arXiv2026

Towards Strategy-Level RSI for Skill-Augmented Agents: Learning When to Reuse Skills from Execution Feedback

Long-running agents accumulate reusable Skills, but a Skill that is semantically relevant to a task is not necessarily worth loading in the current state. We study the applicability question that arises once a candidate Skill is known: should it be loaded in the current state? We propose SkillApt, which uses matched WITH/WITHOUT Skill executions on the same task state as persistent evidence, estimates the conditional marginal utility of the Skill, and chooses LOAD or ABSTAIN accordingly. The base model, agent architecture, and Skill contents stay fixed; only the external deployment policy changes. We call this constrained setting strategy-level recursive self-improvement (Strategy-Level RSI). On 20 Skills and 160 held-out states, as paired evidence accumulates, SkillApt's task success rises from 81.9% under a cold start to 91.3%, matching a strong zero-shot LLM controller; yet SkillApt activates Skills on only 26.3% of states, versus 98.8% for the zero-shot controller. A hard-candidate study shows that non-optimal Skills mostly leave correctness unchanged while raising execution cost, and occasionally cause correctness harm. An ablation shows that a history recording only WITH success makes the policy load almost everywhere, whereas paired evidence substantially improves selectivity. These results indicate that relevance is not applicability: the main effect of execution evidence is not to make the model stronger but to change how existing Skills are deployed, moving the system from near-always loading to selective reuse.

cs.MA↗
arXiv2026

N/O-enhanced chemical enrichment by massive stars: the interplay of optically-thick winds and rotation

Recent JWST observations have revealed a sample of high-$z$ galaxies with highly enhanced N/O abundance ratios, $\log\rm (N/O)\gtrsim -1.1$. Strong optically-thick winds of rotating massive stars going through the Wolf-Rayet (WR) phase provide a viable explanation for such N/O enhancements, supported by the direct spectroscopic WR signatures discovered in a few galaxies. Such strong winds are also needed to reproduce the single WR stars in the Small Magellanic Cloud (SMC). Here, we compute the metal yields of massive ($\sim 20-100\ \rm M_\odot$), metal-poor stars using two sets of MESA single stellar evolution models with and without optically-thick winds in a dense grid of metallicities ($Z_\star\sim 0.001-0.0045$) and initial spins ($ω\equivΩ/Ω_{\rm crit}\sim 0-0.7$). Considering an initial spin distribution based on observations of O-type stars in the SMC, we find that both sets produce strong N/O enhancements up to $\rm \log(N/O)\sim 0.1$ within a few Myr after a starburst. The models with optically-thick winds undergo a rapid transition from high N/O to the `normal' state of $\rm \log(N/O)\sim -1.5$ at $t\sim 3-4$ Myr driven by C/O-rich winds, while the transition is slower and weaker without optically-thick winds. The optically-thick-wind scenario can reproduce the nebular C, N, O, and He abundance patterns of three representative high-$z$ galaxies showing direct WR signatures: RXCJ2248-ID, Sunburst Arc, and MARTA 4327, with stellar metallicities and ages similar to those inferred from SED-fitting. This can be interpreted as a coherent evolution sequence captured before ($t\lesssim 3$ Myr), during ($t\sim 3-4$ Myr), and after ($t\gtrsim 5$ Myr) the transition. In contrast, the optically-thin-wind-only scenario ejects much less C and O via winds and cannot reproduce Sunburst Arc and MARTA 4327. The yields are available at https://zenodo.org/records/22897057.

astro-ph.GA↗
arXiv2026

UniDataAgent: An Ontology-Grounded Agent for Enterprise Question-to-Report Automation

Enterprise data agents must preserve organization specific semantics, not just translate questions into queries. We present ChinaUnicom DataAgent (UniDataAgent), an ontology grounded system for reusable question-to-report analysis that separates semantic acquisition from online execution. Ontology Acquisition and Validation stage (OAV) builds versioned enterprise ontologies from metadata, business knowledge, and supporting materials through expert authored business skills, constrained generation, question verification, and selected expert review. Question-to-Report Execution (QRE) stage retrieves semantic contracts for each question, coordinates skills and data tools, validates results, and produces evidence linked reports. Across 27 enterprise tables and roughly thousands of metric types, ontology construction took a few hours instead of about one week manually. It took just a few minutes to generate the reports, instead of several working days. Ontology grounding achieved 95.0\% strict accuracy on real business questions, versus 72.5\% for document RAG, especially on structured and compositional tasks. The system has already been deployed to generate cost savings and has the potential to be replicated in other enterprises.

cs.CL↗
arXiv2026

Shape without scale: an identifiability dichotomy for a bounded tail observed through a non-additive measurement kernel

A latent severity has a bounded lower tail with density of shape alpha and scale L. It is observed only through a fixed Markov kernel K that is biased and non-additive. The relative conditional spread of K diverges at the endpoint. Our sample is i.i.d. from the marginal Q alone, with no anchoring covariate or instrument. We prove a dichotomy. The shape index alpha is identifiable: for every admissible choice of the class constants, any two observationally equivalent members of a lean class share alpha, determined by a near-endpoint expansion of Q. The rate, namely L and the fixed-scale exceedance p_tau, does not survive. There exist admissible shared class constants and two members of a smaller regularity class whose observed laws coincide exactly. Across the pair alpha agrees, whereas L and p_tau move. A degenerate Le Cam two-point bound excludes any uniformly consistent estimator of either, and pointwise consistency fails at one member. Only the rate needs an anchor. We conjecture that a known kernel family with known edge map identifies the rate fiber by fiber if and only if the family satisfies a fixed-scale injectivity clause, and we prove the sufficiency direction. In surrogate safety, uncalibrated conflict data give the shape of near-crash risk, not its absolute rate.

math.ST↗
arXiv2026

The Tate conjecture for powers of abelian fourfolds and the Hodge conjecture for powers of CM fourfolds

We prove the Tate conjecture in every codimension on every power of an abelian variety of dimension at most four over a finite field. For geometrically simple fourfolds, we use Broe's theorem and the classification of Frobenius relations. For dihedral pairs of abelian surfaces, we construct algebraic classes using families of abelian threefolds, the Gross--Schoen height formula, and polarization contractions. We also prove the Hodge conjecture for every power of a complex CM abelian fourfold, using Markman's theorem on Weil classes and Milne's criterion. The finite field results imply standard conjecture~D and the expected pole orders of the zeta function on every power.

math.AG↗
arXiv2026

Broadband and Large-Multimode Quantum Optical Storage with Rare-Earth Ions in Solids

Broadband quantum memory devices are essential elements for future quantum networks. Here we propose a broadband quantum memory scheme called Hole Anti-hole Grating Echo Memory (HAGEM) for rare-earth ions in solids, which can be implemented with most of the present rare-earth ions hosted in various types of solids. Using HAGEM, a quantum memory device with one billion temporal modes per second can be developed with current available technologies. To validate the memory protocol, we conducted some proof-of-concept experiments. We provide a Eu$^\text{3+}$ molecular complex with special hyperfine level structures of which the hyperfine level separations are in a specific mathematical correlation that can be obtained by harnessing chemical engineering. Using the proposed memory protocol and material, we experimentally demonstrate a single-mode quantum optical storage efficiency of 14.9% and a memory bandwidth of 200MHz, which can easily be extended to a few GHz. In addition, a multi-mode storage of four temporal modes is achieved with a storage efficiency of 1.3%. Broadband & large-multimode quantum memory device with a long storage time enabled by HAGEM integrating with the current available entangled photon pair source and tunable delay lines, a quantum repeater network with an entanglement distribution rate on the order of 1kHz - 10kHz can be created with a communication distance around 100km. This will greatly facilitate the development of quantum networks in near future.

quant-ph↗
arXiv2026

Benchmarking Hyperspectral Foundation Models for Hyperspectral Unmixing

Several foundation models dedicated to hyperspectral images have recently been made available. These models are trained on large unlabeled datasets and exhibit strong performance on many hyperspectral imaging tasks, such as classification or denoising. Nonetheless, their performance for hyperspectral unmixing -- the task of separating mixed spectra of overlapping materials in a hyperspectral image -- remain understudied. This might partly be due to the fact that most of them rely on vision transformer backbones, including patchification, leading to a feature resolution problem. While hyperspectral unmixing already arises from the low resolution of hyperspectral images, this patchification step potentially makes the problem even more ill-posed. Therefore, in this work, we aim to answer two questions: 1) how do foundation models perform in hyperspectral unmixing?; 2) how to tackle the feature-level loss of resolution? To answer the first question, we benchmark foundation models for unmixing, showing that they can reach state-of-the-art performance on four hyperspectral unmixing datasets. To answer the second question, we compare several feature upsampling approaches and empirically show that using a simple one can lead to high performance results. The code is available at https://gitlab.telecom-paris.fr/ring/hfm-hsu.git.

cs.CV↗
arXiv2026

ELF-REG: Scaling Continuous Diffusion Language Models to Reasoning Tasks

Fully continuous diffusion language models (dLMs) denoise continuous representations without intermediate discretization, then decode all response tokens in parallel at the final step. Their performance on challenging reasoning tasks remains less established than that of autoregressive (AR) LLMs and masked dLMs. We scale Embedded Language Flows (ELF) to mathematical reasoning and code generation on GSM8K, MATH-500, HumanEval, and MBPP. We introduce ELF-REG, which improves learning with representation alignment and entanglement (REPA+REG), where a frozen AR teacher supervises intermediate denoiser features and supplies a global representation that is jointly denoised with the response. ELF-REG-L achieves 55.96% pass@1 on GSM8K at 64 network function evaluations (NFE), and 13.39% on MATH-500 and 22.56% on HumanEval at 128 NFE. It outperforms the evaluated comparable-scale dLMs in pass@1 on GSM8K and code, and improves MATH-500 pass@1 from 10.55% for the ELF-L baseline to 13.39% with ELF-REG-L. Without few-step training, the same task-specific checkpoints support strong low-NFE performance through early-stop, which decodes an intermediate clean prediction without completing the denoising trajectory. At 16 NFE, ELF-REG-L reaches 41.21% HumanEval pass@10, outperforming recent continuous dLMs of comparable scale.

cs.CL↗
arXiv2026

PQLS: A Quasilinear Gyrokinetic Transport Solver with a Bayesian Saturation-Rule Closure

Quasilinear models make gyrokinetic turbulent-transport predictions sufficiently fast for integrated modelling, but their predictive capability is limited by two factors: the physical and geometrical applicability of the linear solver, and the validity of the saturation rule used to close the model. We present the Predictive Quasilinear Solver (PQLS), a quasi- linear gyrokinetic transport solver formulated in general magnetic geometry. Its implementation as an eigenvalue solver retains electromagnetic and collisional effects, provides access to dominant and subdominant modes and is differen- tiable with respect to all plasma parameters. Linear benchmarks against GENE reproduce the growth rates, frequencies, and eigenfunctions. We additionally formulate the saturation-rule closure as a Bayesian inference problem that distin- guishes uncertainty in its fitted coefficients from the residual model-form uncertainty. The approach is demonstrated by calibrating the SAT3 rule on PQLS quasilinear weights against published nonlinear CGYRO cases. In addition to improving the robustness of the calibration, the new method also quantifies the uncertainty in each of the fit coefficients. Such uncertainty is propagated through transport calculations to produce error-aware profiles that are compared to the ones obtained from the full gyrokinetic simulation, showing excellent agreement.

physics.plasm-ph↗
arXiv2026

An $n^2\log\log n$ Lower Bound for Permanent Circuits with Valid Division

We record lower bounds for permanent circuits with valid division over characteristic zero, counting nonscalar multiplications and divisions while additions and scalar operations are free. Chapter 5 of OpenAI's Ten Advances supplies the block construction and critical-locus estimate; these, with the parameter choice made here and the classical bounds of Strassen and Baur-Strassen, give liminf as n tends to infinity of L_div(per_n)/(n^2 log_2 log_2 n) >= 1/12. Our finite-parameter refinement for matching-minor polynomials observes that the coefficient vector at deletion size h lies in a space of dimension at most min{binom(t,h), binom(t,d-h)}. Twelve machine-checked declarations of the Lean development accompanying OpenAI's manuscript on border determinantal complexity of the permanent (September 24, 2026) imply, by a short written argument, the existence of a geometric slice; a further written, unformalized application of the same criterion gives L_div(per_n) >= (n^2/119790) log_2(n/44) for n >= 1936. Assuming Proposition 7.3 of that manuscript gives the bound (n^2/86400) log_2(n/48) for n >= 1408. These are order-of-growth results; the elementary Hessian bound n^2/2 is larger at practical orders.

cs.CC↗
arXiv2026

Fibred realizations of the prism quandle $P_7$ in the standard four-sphere

We study smooth fibred two-knots whose fundamental quandle is the prism quandle of order fifty-six. The candidates arise from framed circle surgery along a section of a mapping torus, using the two possible normal-framing classes. We prove that both ambient homotopy four-spheres are diffeomorphic to the standard four-sphere. The candidates have the same knot group and diffeomorphic exteriors, but they are inequivalent even as topological pairs. Every smooth fibred realization of this quandle in a homotopy four-sphere is equivalent, up to knot orientation, to exactly one of these two knots. Standardness follows from a Whitehead-link model for the orbit orbifold together with circle-action recognition. Inequivalence is detected by the third cyclic branched covers, on which a spin-bordism secondary invariant takes different values. Thus the parameter-seven prism quandle admits smooth fibred realizations in the standard four-sphere.

math.GT↗