跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+479新增论文
+361其中 AI 相关
+64已有论文修订
+93新增 AI 导读

arXiv 最近完成扫描:2026-11-16 01:34 · 全站最近入库:2026-11-16 21:56各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 22:20 — 2026-11-16 22:20(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-16 01:34最近发现新增版本:2026-11-16 21:56库内论文最新更新时间:2026-11-14 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

23,827 已收录论文19,941 AI 相关6,922 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 17,708 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33(北京时间)
新论文 · v1arXiv:2611.15436

AgentTorrent: A Dataset of Context and Harness Engineering Artifacts on GitHub

Sebastian Baltes, Seyedmoein Mohsenimofidi, Levi Böhme, Vahram Nersesyan, Muhammad Auwal Abubakar, Jai Lal Lulla, Matthias Galster, Christoph Treude

原文摘要

Over the past months, developers have moved from formulating detailed prompts to engineering the context that agentic coding tools read and to configuring the underlying harnesses. For example, developers commit context files, skills, hooks, and MCP configurations to their repositories, and these files control which instructions a harness reads and which actions it can take. Developers may also commit the specs, plans, and task files that agentic coding tools and spec-driven development frameworks write for agents to implement. The AgentTorrent dataset records these files for six proprietary and open-source agentic coding tools in 10,944 engineered software projects on GitHub. It contains 62,374 configuration artifacts, the files that these artifacts import, 27,593 agentic planning documents, and more than 1.2 million commits that credit an AI tool. All heuristics used for data collection are grounded in tool documentation or derived from GitHub repositories, and we evaluate their detection accuracy with each collection. We use open-weight models to classify whether GitHub repositories host engineered software projects. Unlike most datasets, AgentTorrent links configuration artifacts, imported files, agentic planning documents, and commits that credit an AI tool within the same repositories. Configuration artifacts, planning documents, and credited commits occur together in 388 repositories. For example, our dataset allows researchers to study whether a tool is credited in commits that add instructions for that same tool and how instruction files import and contradict each other. Our data collection pipeline supports regular updates and automatically checks vendor documentation and reports unknown artifacts and agentic coding tools. The dataset and the pipeline are open-source and are archived on Zenodo.

cs.SE

新论文 · v1arXiv:2611.15435

ThermalSplat: Generalizable Pose-Free Gaussian Splatting for Sparse-View Thermal 3D Reconstruction

Prasanna S, Gopi Raju Matta, Divya Madhuri Vemunuri, Kaushik Mitra

原文摘要

Thermal infrared imaging is vital for applications like search-and-rescue and autonomous navigation, yet trans- lating it into 3D scene representations remains a critical bottleneck. The low-texture nature and view-dependent emissivity of thermal images cause standard pose estima- tors to fail catastrophically, collapsing standard 3D Gaus- sian Splatting pipelines under sparse-view constraints. Fur- thermore, existing thermal reconstruction methods rely heavily on dense capture and computationally expensive per-scene optimization. Critically, many of these ap- proaches depend on aligned RGB-thermal datasets to ob- tain reliable camera poses and initial point clouds, ren- dering them unusable in purely thermal, unposed envi- ronments. To break this bottleneck, we present Thermal- Splat, the first generalizable, feed-forward Gaussian splat- ting framework that reconstructs 3D scenes directly from sparse thermal images without requiring known camera poses, RGB guidance, or per-scene training. We over- come thermal-specific challenges using a soft census loss for brightness-invariant structural matching and a bound- ary structure loss to explicitly supervise temperature discon- tinuities. Trained on an aggregated multi-source dataset, ThermalSplat synthesizes sharp reconstructions of unseen scenes in a single feed-forward pass. Extensive experi- ments demonstrate that ThermalSplat holds a strict advan- tage over scene-specific thermal splatting methods in the sparse regime. Furthermore, even compared against these baselines operating under their ideal dense capture set- tings, ThermalSplat achieves superior perceptual fidelity. By eliminating the need for computationally expensive per scene optimization, ThermalSplat offers an efficient and generalizable approach to thermal 3D reconstruction.

cs.CV

版本更新 · v2arXiv:2611.12680

Wolfe's Minimum-Norm-Point Method Generalizes the Simplex Method

Yann Disser, Nils Mosis

原文摘要

Wolfe's method computes the point of minimum Euclidean norm in the convex hull of finitely many points. The method is parameterized by an insertion rule and it has been an open problem whether Wolfe's original rule guarantees a polynomial number of iterations (De Loera, Haddock, and Rademacher; STOC'18). We resolve this question by proving an exponential lower bound. In fact, we establish a tight connection to the simplex method for linear programming that allows us to transfer many lower bounds for simplex pivot rules to corresponding insertion rules of Wolfe's method. These include not only Wolfe's linopt rule via Dantzig's pivot rule, but also many other classical, randomized and history-based rules. As a corollary, we obtain the first exponential example for the Lawson-Hanson algorithm for nonnegative least squares, as well as PSPACE-hardness of deciding whether the linopt rule ever uses a given point.

math.OC · cs.DS

新论文 · v1arXiv:2611.15434

SurvEval: Listwise Re-Ranking Evaluator via Survival-Based Exposure and Multi-Objective Value Modeling

YaChen Yan, Liubo Li

原文摘要

Re-ranking is a crucial stage in recommender systems, yet existing evaluators oversimplify user browsing assumptions and collapse multi-objective signals into a single scalar, limiting their ability to reflect true listwise utility. We propose \textbf{SurvEval}, a listwise evaluator that jointly models user exposure and item value for re-ranking evaluation. We formalize the sequential browsing process as a discrete-time survival model, deriving monotonically decreasing reaching probabilities that capture the cascading structure of user browsing. We further introduce a multi-objective value estimation module with dedicated per-objective prediction heads and uncertainty-based loss weighting to adaptively balance objectives of varying scales and noise levels. Both components are jointly optimized via a unified transformer backbone, and the listwise utility is obtained by summing the position-wise products of reaching probabilities and conditional item value estimates. Extensive experiments on KuaiRand and Avito demonstrate that SurvEval achieves superior performance over state-of-the-art baselines.

cs.IR

新论文 · v1arXiv:2611.15433

Verifier Choice is a Benchmark Design Variable: Auditing Structural Counting Evaluation in Text-to-Image Models

Shurun Li, Michael Hua Wang, Haibo Zeng, Long Wang

原文摘要

Automatic evaluation is now central to text-to-image benchmarking, but the verifier that converts generated images into scores is often treated as an implementation detail. We argue that this is unsafe: an automatic benchmark score is a property of the image generator, prompt distribution, verifier, verifier-use protocol, and scoring rule, not of the generator alone. We study this issue in structural counting, a controlled setting where human-visible counts can be audited. We construct a verifier-centric audit suite over seven counting objects: cell phones, chairs, and traffic lights for object counting, and clover leaflets, dresser drawer fronts, power-strip sockets, and turbine blades for substructure counting. Across 6,240 generated images from Qwen-Image-2512 and FLUX.2-dev, we compare open MLLM verifiers under count-only, reject-aware, zero-shot and few-shot protocols, plus TIFA-like target queries, a detector baseline, and frontier closed-model checks on two hard objects. On a balanced 2,280-image evaluation split, aggregate score matching can coexist with image-level disagreement, and verifier-use protocols and generator distributions change the measured reliability. Every verifier-protocol configuration also shrinks the human score difference between the two generators, by 7.5-25.0 pp, and with a third generator, Z-Image, all four tested protocols reverse the human ordering of Z-Image and FLUX.2-dev on our fixed prompt suite. GPT-5.5 and Claude Opus 4.7 checks do not remove the failure modes: GPT-5.5 still has only 59.92% clover valid-count fidelity, while Claude Opus 4.7 has lower same-slice agreement ($κ=0.432$). We propose reporting automatic benchmark scores with a factorized verifier reliability profile: aggregate score gap, image-level agreement, generator-comparison distortion, valid-image count fidelity, rejection behavior, output compliance, and directional bias.

cs.CV · cs.LG

版本更新 · v2arXiv:2611.02275

Data-Scarce, Multi-Granular, Multimodal: Benchmark Design for Biosynthetic Gene Clusters

Hanlin Xiao, Eriko Takano, Mauricio A. Álvarez, Rainer Breitling

原文摘要

Biosynthetic gene clusters (BGCs) are co-located genes that encode the biosynthesis of specialized metabolites, an important source of antibiotics, anticancer agents, immunosuppressants, and agricultural bioactives. Deep learning, particularly self-supervised foundation models, has spurred interest in BGC sequence modeling, but evaluation has lagged behind: existing benchmarks often rely on coarse classification over small curated datasets, providing limited discrimination among models and insight into downstream utility. We propose a benchmark-design framework in which an informative benchmark should be discriminative, separating models with statistical confidence, and applicable, spanning scientifically meaningful tasks under different levels of generalization. We instantiate this framework as BGC-Bench, which derives five tasks from MIBiG, ranging from A-domain substrate prediction to BGC-molecule retrieval, evaluates them under genotype- and phenotype-shift splits, includes an active-learning setting for label-scarce discovery, and is easily extensible to new tasks and encoders. BGCs further provide a shared testbed across granularities and modalities: BGC-Bench compares twelve encoders on nucleotide, amino-acid, or Pfam-domain sequences, including compositional baselines for each input type, and combines BGC and chemical encoders for cross-modal retrieval. No single representation dominates: the best granularity depends on the biological scale of the task, and pretrained encoders do not consistently outperform other representations across tasks and splits. Active-learning experiments show that most tasks remain data-limited, encoder rankings change with labeling budget, and acquisition strategies yield modest gains. These results show that BGC model evaluation requires task-diverse, shift-aware benchmarks, and the framework can be generalized to benchmark design beyond BGCs.

cs.AI

新论文 · v1arXiv:2611.15431

Subadditivity and the failure of strong subadditivity in the Nichols--Wootters family of entropies

Lin Zhang, Wenlin Zhang, Haobo Ma

原文摘要

We prove subadditivity for the Nichols--Wootters family of entropies interpolating between the von Neumann entropy and the subentropy, with all spectra evaluated in one common ambient dimension. Beyond the von Neumann entropy, equality holds exactly when at least one marginal is pure. This settles the subadditivity conjecture concerning subentropy. Every member except the von Neumann entropy fails strong subadditivity, already on a maximally entangled two-qubit state tensored with a mixed qubit. Its strong-subadditivity defect takes both signs on quantum Markov states. For a pure bipartite state tensored with an ancillary state, we identify the defect with a separately concave product gap and determine its exact maximum under rank constraints. For arbitrary bipartite states, we obtain an explicit lower bound in the marginal purities. For two-qubit states, we determine the optimal coefficient multiplying the product of the smaller marginal eigenvalues in every common ambient dimension, including the subentropy coefficient $1/3$. A positive representation by Haar averages of compressed matrix entropies gives the pinching inequality underlying these results and also yields entropy-gain bounds for unital quantum channels.As an application, each entropy equals the classical communication capacity retained after a specified isotropic partial projective measurement, at a prescribed average input state, even when entangled block encodings are allowed.

quant-ph · cs.IT · math-ph

新论文 · v1arXiv:2611.15427

Overcoming the Resolution Limit: Significance-Aware Regularization for Intersectional Fairness

Antonio Ferrara, André Panisson, Francesco Cozzi, Alan Perotti, Francesco Bonchi

原文摘要

Intersectional fairness requires evaluating models across fine-grained subgroups. However, as group definitions become more granular, the sample size per subgroup shrinks, causing fairness estimates to become increasingly unstable. Existing in-processing methods are certainty-blind: they penalize observed disparities without distinguishing statistically certified unfairness from sampling noise. This leads to fairness overfitting, where models sacrifice utility to "fix" spurious disparities in sparse regimes. To address this, we introduce the significant-excess metric family, a continuous, uncertainty-aware fairness score that measures only the portion of a subgroup's disparity exceeding what sampling variation can explain at a specific confidence level. Building on this metric, we propose SAFER (Significance-Aware Fairness Regularizer), a differentiable in-processing method that gates fairness penalties using a soft-count likelihood-ratio evidence statistic. SAFER satisfies a provable safety property: when every subgroup disparity lies well below the gate threshold, the unconstrained optimum is nearly stationary for SAFER. At that optimum, the back-propagated contribution of its significance-gated component is exponentially small, and the additional fairness contribution comes from a small ungated term with a fixed weight. Mechanism ablations across three benchmark datasets and purpose-built stress variants show that removing any component of SAFER (threshold smoothing, the significance gate, or a dense-tail term) produces the failure the theory predicts on at least one stress variant. In sparse intersectional settings, SAFER achieves a better fairness-utility Pareto frontier than the baseline methods, and it preserves utility when certified unfairness is absent.

cs.LG

新论文 · v1arXiv:2611.15426

SimRec: Learning to Search, Ask, and Recommend from Simulated Interaction

Yaxin Zhang, Xingyu Wu, Yuchen Yan, Yuqin Dai, Hong Zhao, Lixing Shen, Zhan Chen, Jian Shao 等 9 位作者

原文摘要

LLM-based recommendation agents can search catalogs, inspect products, ask users, and revise rejected recommendations. However, these behaviors are often governed by hand-designed workflows, while existing training methods optimize isolated components such as query rewriting or list ranking rather than the interaction policy as a whole. We present SimRec, a reinforcement learning framework that trains the full interaction policy of a tool-integrated recommendation agent. SimRec introduces two components. A task-anchored user simulator provides feedback on every recommendation: conditioned on the current task need rather than on user profile, and it can be an external LLM or the recommender itself under self-play. A task-aligned reward optimizes the whole trajectory with GRPO, crediting natural retrieval quality and valid tool use alongside the final hit, with training-only target injection to handle sparse hits over large catalogs. On ESCI and Amazon C4, SimRec improves the NDCG@100 of a Qwen3-4B agent from 32.23% to 35.42% and from 22.94% to 24.72%}, outperforming training-free agents on far larger backbones and task-aligned baselines on NDCG. Analysis shows that the trained agent verifies candidates, uses preferences, and recovers from rejected recommendations far more often than its untrained counterpart. Our code is available at https://github.com/ZJU-REAL/SimRec .

cs.CL · cs.AI · cs.IR

新论文 · v1arXiv:2611.15425

A polynomial-time approximation scheme for the weighted Euclidean Minimum Latency Problem in fixed dimension

Daniel Skachkov, Yuriy Dorn

原文摘要

In the Traveling Salesman Problem, a route must visit all customers and return to its starting point while minimizing its total length. The Minimum Latency Problem, also known as the Traveling Repairman Problem, instead minimizes the sum of the times at which customers are first visited. We give a deterministic polynomial-time approximation scheme for the weighted Euclidean Minimum Latency Problem in every fixed dimension, with arbitrary nonnegative customer weights. Our proof restricts the form of a solution in three steps, each with a controlled increase in cost. First, a decomposition based on Sitters' method assigns customers to phases that start and end at the root. Second, within each phase, we divide the route into a constant number of blocks and charge each customer at the end of its assigned block. The resulting objective is a nonnegative linear combination of block lengths. Third, we add an auxiliary dimension and use a fixed Minkowski norm to close each block path into a tour. Arora's closed-tour estimate then gives paths that cross the boundaries of a common dissection only at designated portals and a bounded number of times per face. A dynamic program optimizes the restricted objective, and projecting its paths back to the original space gives the final route. Rounding weights and coordinates keeps the state space polynomial.

cs.DS · math.CO

版本更新 · v2arXiv:2602.06594

Machine Learning Practitioners' Views on Data Quality in Light of EU Regulatory Requirements: A European Online Survey

Yichun Wang, Kristina Irion, Paul Groth, Hazar Harmouch

原文摘要

Understanding how data quality aligns with regulatory requirements in machine learning (ML) systems presents a critical challenge for practitioners navigating the evolving EU regulatory landscape. To address this, we first propose a practical framework aligning established data quality dimensions with specific EU regulatory requirements. Second, we conducted a comprehensive online survey with over 180 EU-based data practitioners, investigating their approaches, key challenges, and unmet needs when ensuring data quality in ML systems that align with regulatory requirements. Our findings highlight crucial gaps between current practices and regulatory expectations, underscoring practitioners' need for more integrated data quality tools and better collaboration between technical and legal practitioners. These insights inform recommendations for bridging technical expertise and regulatory compliance, ultimately fostering responsible and trustworthy ML deployments.

cs.DB

版本更新 · v3arXiv:2604.03641

Delayed homomorphic reinforcement learning for environments with delayed feedback

Jongsoo Lee, Jangwon Kim, Soohee Han

原文摘要

Reinforcement learning in real-world systems often involves delayed feedback, which violates the Markov assumption and impedes both learning and control. Canonical augmentation-based approaches address this issue by augmenting the state with action histories, but suffer from state-space explosion and the resulting sample-complexity burden. Despite recent progress, state-of-the-art augmentation-based methods either primarily alleviate this burden for the critic or employ separate mechanisms for the actor and critic. To address this limitation, we propose delayed homomorphic reinforcement learning (DHRL), a framework grounded in MDP homomorphisms that introduces a belief-equivalence relation over the augmented state space to collapse control-redundant augmented states into a compact abstract state space. This framework yields exact abstraction under deterministic dynamics and approximate abstraction under stochastic dynamics, providing a unified abstraction mechanism that benefits both the actor and critic. In finite domains, exact abstraction preserves optimality and recovers the delay-free sample-complexity order, while approximate abstraction admits a bound on the value loss of the resulting policy. For continuous domains, we introduce deep delayed homomorphic policy gradient (D$^2$HPG), a deep actor-critic instantiation of the DHRL framework. Experiments on MuJoCo continuous-control benchmarks demonstrate that D$^2$HPG outperforms strong augmentation-based baselines, especially in long-delay regimes. Code is available at https://github.com/JongsooLee/DHRL.

cs.LG · cs.AI

版本更新 · v3arXiv:2603.11703

EvoFlows: Evolutionary Edit-Based Flow-Matching for Protein Engineering

Nicolas Deutschmann, Constance Ferragu, Jonathan D. Ziegler, Shayan Aziznejad, Eli Bixby

原文摘要

Protein optimization improves an existing protein through targeted mutations, but existing protein language models are poorly suited to it: autoregressive models require full sequence generation, masked language and discrete diffusion models rely on pre-specified mutation locations, and none naturally support insertions and deletions relative to a template. We introduce EvoFlows, a variable-length sequence-to-sequence approach that learns mutational trajectories between evolutionarily related proteins via edit flows, predicting not only _which_ mutation to perform, but also _where_ and _how many_ should occur. Across diverse protein families from UniRef and OAS, EvoFlows generates variants that remain consistent with natural families while exploring farther from the template than leading baselines. _In vitro_ validation of a VHH binder and an enzyme confirms that the large majority of generated variants remain functional and stable without computational downselection.

cs.LG

新论文 · v1arXiv:2611.15420

Learning Motion-Appearance Coupling Priors for Solving Video Inverse Problems

Anselm Krainovic, Reinhard Heckel

原文摘要

Reconstructing video from degraded measurements requires priors that capture both frame appearance and temporal dynamics. Many recent approaches combine pretrained image diffusion priors with motion guidance, sidestepping the costs of training large video models. Temporal consistency is often enforced via photometric losses or warping-based constraints, which implicitly assume that motion explains frame-to-frame or latent-to-latent changes up to small deviations, an assumption that fails under occlusions, lighting changes, and non-rigid deformation. In this paper, we propose to learn the coupling between motion and frame appearance as a generative prior. We learn the distribution of motion-induced appearance residuals with a conditional diffusion model and use it to regularize reconstruction in video inverse problems. Our motion-appearance prior acts as a plug-in regularizer for diffusion-based video reconstruction, preserving the benefits of image diffusion models while remaining compatible with existing appearance regularization techniques. Experiments on dynamic video inverse problems show that reconstruction with a learnt motion-appearance coupling prior outperforms photometric-consistency and noise-warping baselines, and that the approach also transfers to cardiac MRI reconstruction.

cs.CV · cs.LG · eess.IV

版本更新 · v2arXiv:2611.07535

Characterizing CUDA Locality Domains on Blackwell GPUs

Rocky Wang, Joyce Zhou, Keith Jiang, Jason Zhao

原文摘要

Low latency is critical to user experience. As NVIDIA Blackwell and Rubin GPUs advance multi-die scaling, data placement and cross-die communication become increasingly important. However, the latency implications and underlying costs of cross-die communication and die-aware execution remain largely unexplored. In this report, we present a comprehensive evaluation of the Locality Domain (LD) feature introduced in CUDA 13.4. Conducted on a single node equipped with eight NVIDIA B300 GPUs, our study examines three aspects: memory management and access, computation, and communication, to identify the sources of latency and, more importantly, explore how LD can mitigate them. Our measurements show that the benefits of LD depend on data access patterns. Through die-aware adaptations and integration with widely used libraries and tools, we achieve speedups of up to 43.5% for NCCL 2.32.3 (AllGather), 111.3% for nvbandwidth (DtoD), and 37% for DeepGEMM (EP). We also identify non-negligible LD-related GPU memory allocation overhead when allocation sizes exceed 32 MB due to barrier and TLB invalidation, we propose a fix to the GPU driver kernel module (submitted to NVIDIA) and show that overhead could be reduced 7x ~ 10x without requiring any user-code changes.

cs.DC

版本更新 · v2arXiv:2611.06998

Resolving Tusnády's Problem

Nicola Wengiel

原文摘要

We prove a lower bound of $Ω_d((\log N)^{d-1/2})$ for the worst-case combinatorial discrepancy of $N$-point sets with respect to anchored boxes in every fixed dimension $d\ge3$. Together with Nikolov's upper bound of the same order and Wei's planar lower bound, this settles Tusnády's problem: the discrepancy of axis-parallel boxes is $Θ_d((\log N)^{d-1/2})$ in every fixed dimension $d\ge2$. The proof combines smoothed product test functions, sampling discretization, and an entropy argument for conditioned Gaussian vectors. The matching upper and lower bounds are also formalized in Lean 4.

math.CA · cs.CG · cs.DS · math.CO

版本更新 · v2arXiv:2611.14371

Reliable and explainable analysis of AVS/RS by Statistical Model Checking and Process Mining

Emanuele Guerrazzi, Andrea Vandin

原文摘要

Automated Vehicle Storage and Retrieval Systems (AVS/RS) are increasingly adopted to boost warehouse throughput while containing operational costs. We present a tool-supported analysis workflow that combines discrete-event simulation, Statistical Model Checking (SMC), and Process Mining (PM) techniques to quantify and explain AVS/RS performance. SMC, implemented through the MultiVeStA tool, delivers statistically reliable estimates of key indicators, such as cycle times and inventory consolidation, across alternative stocking and picking policies with user-specified confidence levels. MultiVeStA orchestrates multiple independent simulations in a black-box manner, enabling automated verification. PM ingests the resulting event logs to uncover the behavioural patterns and bottlenecks that drive those indicators, closing the interpretability gap typical of simulation studies. We instantiate the framework on a new open-source AVS/RS simulator configured according to recent industrial guidelines and evaluate multiple control strategies and different scenarios. Results show that the coordinated use of SMC and PM supports both predictive what-if analyses and post-hoc diagnostics. To the best of our knowledge, this is the first application of a combined SMC-PM workflow to AVS/RS operations, complemented by a reusable open-source simulation and analysis toolchain.

cs.SE · eess.SY

新论文 · v1arXiv:2611.15413

MORA: Correcting Time Series Forecasts with Coarse Future Signals

Hu Li, Kenan Li, Haonan Xu, longlong, Shuqin Gao, Yucheng Zhang, Feng Dai

原文摘要

Recent architectural advances have improved time-series forecasting. However, our analysis of forecasters with diverse architectures reveals a common pattern: a substantial share of residual energy is concentrated in a small number of low-frequency components. Simple linear probes further show that the forecasters' historical inputs still contain predictive information about errors in the forecast mean. Building on these findings, we propose Model-Output Residual Adaptation (MORA), a model-agnostic residual adapter that refines the outputs of frozen forecasters without requiring additional inputs. MORA estimates segment-wise means of the target sequence through structural extrapolation and learned estimation, rather than directly predicting backbone-specific residuals. It then converts discrepancies between these estimates and the corresponding statistics of the backbone forecast into stagewise corrections, scaling each update according to estimation reliability. Extensive experiments on twelve benchmarks with six forecasting backbones show average MSE reductions of 7.2% for long-term forecasting and 28.1% for short-term forecasting, with lower MSE in 96.2% of the 288 evaluated cases. The benefits also extend to three time-series foundation models evaluated at different context lengths. Together, these results demonstrate that predictable coarse-scale residuals can be effectively exploited to improve existing forecasts. Code repository: https://anonymous.4open.science/r/MORA-B0B0/.

cs.LG

版本更新 · v2arXiv:2504.15779

Shannon observables: A scalable approach to information decomposition

Aaron J. Gutknecht, Fernando E. Rosas, David A. Ehrlich, Abdullah Makkeh, Pedro A. M. Mediano, Michael Wibral

原文摘要

Distributed systems, such as biological and artificial neural networks, process information via complex interactions engaging multiple subsystems, resulting in high-order structure of the information within the system with distinct properties across scales. Investigating how these systems process information remains challenging due to difficulties in defining appropriate multivariate metrics and ensuring their scalability to large systems. To address these challenges, we introduce a novel framework based on what we call "Shannon observables": quantities that capture essential properties of high-order information processing, yet can be computed directly from Shannon entropies and scale efficiently to large systems. Our theoretical results demonstrate how Shannon observables can be used to resolve long-standing ambiguities regarding the interpretation of widely used multivariate information-theoretic measures. On the practical side, we demonstrate the framework's versatility across several domains. An analytical study of error-correcting codes shows how Shannon observables trace out distinct encoding regimes as channel noise varies; analyses of deep classifiers and convolutional autoencoders reveal characteristic information-processing signatures across layers and over training; and applied to magnetoencephalography recordings from the human brain, the framework yields a cortical atlas exposing pronounced regional structure in directed information transfer. Overall, our framework resolves fundamental limitations in analysing high-order information phenomena and offers broad opportunities for theoretical developments and empirical analyses.

cs.IT · cs.AI · cs.LG · nlin.AO · physics.data-an

版本更新 · v2arXiv:2601.13867

Block-Fitness Modeling of the Global Air Mobility Network

Giulia Fischetti, Anna Mancini, Giulio Cimini, Jessica T. Davis, Abby Leung, Alessandro Vespignani, Guido Caldarelli

原文摘要

Accurate representations of the World Air Transportation Network (WAN) are fundamental inputs to models of global mobility, epidemic risk, and infrastructure planning. However, high-resolution, real-time data on the WAN are largely commercial and proprietary, therefore often inaccessible to the research community. Here we introduce a generative model of the WAN that treats air travel as a stochastic process within a maximum-entropy framework. The model uses airport-level passenger flows to probabilistically generate connections while preserving traffic volumes across geographic regions. The resulting reconstructed networks reproduce key structural properties of the WAN and enable simulations of dynamic spreading that closely match those obtained using the real network. Our approach provides a scalable, interpretable, and computationally efficient framework for forecasting and policy design in global mobility systems.

physics.soc-ph · cs.SI

新论文 · v1arXiv:2611.15408

AraCrypto-ML: A Multi-Label Corpus of Communicative Functions in Arabic Facebook Cryptocurrency Discourse

Mabrouka Bessghaier, Mohamed Rashad Rouine, Md. Rafiul Biswas, Youssef Ayadi, Wajdi Zaghouani

原文摘要

Arabic cryptocurrency discourse on social media is communicatively heterogeneous: a single post may report prices, explain protocols, promote investment, warn of scams, or invoke Islamic jurisprudence. Yet no annotated resource captures this functional diversity. We introduce AraCrypto-ML, a 10,000-post multi-label corpus of Arabic Facebook cryptocurrency discourse (2014-2024) from 2,898 public pages using a set of related keywords. Collected posts are annotated into six categories: Market Commentary, Educational, Promotional, Skeptical-Warning, Religious Framing, and Other/Unclear, using Fanar-2-27B under a prompt-only multi-label protocol. We validate 1,000 posts against adjudicated gold labels (macro kappa = 0.74) and apply Rogan-Gladen calibration to correct labelling bias. The two annotators agree with each other at macro kappa = 0.62 before adjudication, below the model's agreement with either of them (0.70 and 0.69). Reaction profiles differ strongly by function: Love is about 70% of reactions per post on Educational and Promotional posts but about 33% on Skeptical-Warning, which draws the largest Haha share (about 39%).

cs.CL

新论文 · v1arXiv:2611.15404

Is the Importance Ratio Necessary for Stable Reinforcement Learning in LLMs?

Shuibai Zhang, Junhyuck Kim, Gyeongman Kim, Jaewoong Cho

原文摘要

Reinforcement learning (RL) has become central to post-training large language models (LLMs). However, popular RL methods like GRPO incur non-negligible overhead by computing both old-policy and current-policy likelihoods to form importance sampling ratios. In this work, we propose Likelihood-Gated Policy Optimization (LGPO), which enforces a soft trust region constraint via likelihood-based gating, eliminating the need to compute old-policy likelihoods. Empirically, we show that removing the importance sampling correction term does not harm training stability, whereas removing the trust region mechanism can lead to collapse. Moreover, ratio-based clipping can fail in fully on-policy training: the importance ratio stays at 1, so the ratio-based trust region constraint never activates. Under standard training settings where GRPO is stable, LGPO achieves comparable training stability and peak performance while reducing training time by ~18% on average relative to a standard implementation that recomputes old-policy likelihoods. In fully on-policy training, where GRPO fails, LGPO remains stable, enabling more efficient and robust LLM RL post-training across training regimes.

cs.LG · cs.AI · cs.CL

版本更新 · v2arXiv:2602.14086

Neural Optimal Transport in Hilbert Spaces

Jae-Hwan Choi, Jiwoo Yoon, Dohyun Kwon, Jaewoong Choi

原文摘要

Functional data such as time series, trajectories, and fields are increasingly central in machine learning and are naturally modeled in infinite-dimensional Hilbert spaces. We study Semi-dual Neural Optimal Transport (SNOT) for learning transport maps between probability measures on separable Hilbert spaces. A critical challenge in this setting is the spurious solution problem: a neural map can globally optimize the SNOT objective while failing to recover an optimal transport map. We show that this failure arises from non-regular source measures, which make the inner minimization in the semi-dual objective non-identifiable. We prove that regular source measures, defined via Gaussian null sets, restore inner-minimizer uniqueness and recover the Monge map. For singular sources, we introduce Gaussian smoothing and establish a necessary and sufficient kernel condition for smoothing to restore regularity. We further prove plan-level consistency: as smoothing vanishes, every Wasserstein accumulation point of the induced optimal plans solves the original Kantorovich problem. Experiments on synthetic singular transports and real-world unpaired time-series imputation benchmarks validate the theory, achieving the best performance on several benchmarks and competitive performance on the remaining ones.

cs.LG

版本更新 · v3arXiv:2412.03395

An even simpler hard variant of Not-All-Equal 3-SAT

Andreas Darmann, Janosch Döcker, Britta Dorn

原文摘要

We investigate Positive Not-All-Equal 3-Sat under a strong structural restriction and delineate a sharp tractability threshold. In particular, we consider the restriction of Positive Not-All-Equal 3-Sat to instances that simultaneously satisfy the following properties: (i) the clauses can be decomposed into a disjoint union of $k$ partitions of the variable set, and (ii) the collection of clauses is linear, that is, each pair of distinct clauses shares at most one variable. We show that, under these conditions, Positive Not-All-Equal 3-Sat remains NP-complete for any fixed $k \geq 4$. On the positive side, we prove that for $k \leq 3$, the problem is in P and, notably, all such instances are not-all-equal-satisfiable. Our results do not only show robustness of NP-completeness despite the strong structure imposed but also have implications for hypergraph coloring. In particular, by the well-known correspondence between Positive Not-All-Equal 3-Sat and hypergraph coloring, we obtain the following corollary of our results: For $k\geq 4$, Bicolorability is NP-complete for linear 3-uniform $k$-regular hypergraphs even if the edges are given as a decomposition into $k$ perfect matchings; with the same restrictions, for $k \leq 3$ Bicolorability is in P, and all such hypergraphs are bicolorable. In addition, we show that loosening the restriction to 3-clauses by also allowing 2-clauses shifts the above-mentioned tractability threshold. Moreover, we deduce that every instance of Positive Planar Not-All-Equal Sat with at least three distinct variables per clause is nae-satisfiable. Besides, for Linear Not-All-Equal 3-Sat we derive a computational complexity dichotomy with respect to the number of variable appearances.

cs.CC

新论文 · v1arXiv:2611.15394

Tracking What Works: Social Engineering Lures as Signatures for Victim-Centric Infostealer Campaign Discovery

Estelle Ruellan

原文摘要

Despite advances in antivirus technology, infostealer infections persist at scale, driven primarily by social engineering rather than exploits. Yet, existing campaign detection efforts have focused on technical artifacts, overlooking the social-engineering lures exploiting human behavior and driving successful infections. To address this gap, we introduce a dual-framework system that discovers and tracks infostealer campaigns using victim-side artifacts from stealer logs. By treating social engineering lures as campaign signatures, the framework automatically aggregates individual infections into high-fidelity campaigns in near real-time. Recognizing distinct patterns between software and gaming ecosystems, we developed ecosystem-specific pipelines detecting all large and medium campaigns in both ecosystems (one medium gaming campaign was only recovered merged with another), though performance decreased for small campaigns. Analysis of 177 real-world campaigns identified by our framework reveals distinct temporal and infrastructure patterns with direct remediation implications. Software campaigns accumulate over 60% of infections within the first ten days, demanding rapid intervention following detection, while gaming campaigns grow more intense the longer they last, requiring continuous monitoring. Using social engineering lures as campaign signatures provides robustness to infrastructure fragmentation where traditional IoC-based detection fails. This work shifts campaign tracking from infrastructure correlation to victim-centric analysis, capturing why campaigns succeed rather than only how they operate, enabling awareness-driven prevention that addresses the human decision-making dimension absent from existing research.

cs.SI · cs.CY

新论文 · v1arXiv:2611.15393

VR-Net: Subarray-Level Inter-Cell Coordination for Multi-Cell XL-MIMO under Near-Field Spatial Non-Stationarity

Muhammad Shoaib Ayub, Rana Muhammad Subhan, Insoo Koo, Felipe Augusto Pereira de Figueiredo

原文摘要

Spatial non-stationarity in multi-cell extremely large-scale MIMO (XL-MIMO) confines each user to a visibility region (VR) of each array. VR-Net exploits these regions for subarray-level coordinated beamforming. A base-station (BS)-indexed conflict graph identifies overlapping VRs; gain-based association and a masked multi-cell weighted minimum mean square error (WMMSE) iteration restrict beam design and channel state information (CSI) uploads to the relevant subarray blocks. We establish conditions for exact per-cell decoupling, derive the expected CSI loads, bound the loss of designing on the VR-gated channel and the loss of the support constraint, and characterize single-edge visibility contrast through a closed-form Fresnel transition width. Simulations use a 3GPP TR 38.901 Release-19-based near-field UMi channel with modular facade arrays and pre-registered criteria. At a 30 dB threshold, VR-Net retains at least 99.6% of the full-aperture WMMSE sum rate in every tested setting while uploading 10-39% fewer complex CSI entries. It gains up to 3.5 bit/s/Hz over an all-user adaptation of the multi-cell scheme of Pérez-Adán et al. at moderate power and dense load, and inter-cell coordination adds up to 33% over per-cell design.

cs.IT · eess.SP · math.ST

新论文 · v1arXiv:2611.15392

Integrality Gap of 3.4 for Santa Claus: A Topology-based Analysis

Kimon Boehmer, Chien-Chung Huang

原文摘要

In the \textsc{Max-Min Allocation} problem, a set of indivisible items are to be assigned to a set of players and the goal is to maximize the minimum value among all players. A much studied special case, the \textsc{Santa Claus} problem, assumes that an item has a fixed utility but a player is only interested in a subset of items. The main technique for designing approximation algorithms for \textsc{Santa Claus} is the configuration LP, introduced by the Bansal and Sviridenko. A long line of research has progressively improved the upper bound of the integrality gap, mostly relying on the combinatorial arguments of Haxell for finding perfect matchings in certain hypergraphs. Remarkably, a recent paper of Haxell and Szabó (SODA '23) introduced a novel topology-based analysis, by which they are able to obtain the current best upper bound of $3.534$. In this work, we adapt the approach of Haxell and Szabó to analyze the integrality gap. The main new ingredient in our adaptation is that we stop the process of "edge explosions" in a graph used by Haxell and Szabó when it becomes too costly. This early termination of explosions, combined with a new lower bound on a topological connectivity parameter based on a theorem of Aharoni, Berger, and Meshulam, allows us to improve the integrality gap to $3.4$. Our analysis has the additional advantage of being simpler.

cs.DS · cs.DM

新论文 · v1arXiv:2611.15389

A physics-constrained deep reinforcement learning method for autonomous mission planning of Titan deep space probes

Hanmin Li, Tao Wang, Shuangliang Liu, Zhe Zhang, Maodeng Li

原文摘要

Titan is a compelling target for future deep space exploration, while its extreme Earth-Titan distance and complex environment make continuous ground-in-the-loop planning impractical. Existing planning methods, primarily developed for Earth-observation scenarios, have limited generalizability to uncertain deep space conditions and adaptability to time-varying mission states. To address these challenges, this paper proposes a physics-constrained maskable proximal policy optimization (PCM-PPO) method for Titan autonomous mission planning. A Titan-specific discrete-time Markov decision process is established to characterize mission feasibility and resource constraints. A system dynamics and geometric model is incorporated to enhance physical representation, including central gravity, J2 perturbation, third-body attraction, and downlink occultation. To accommodate opportunity targets under a bounded action dimension, PCM-PPO organizes predefined and opportunity targets through a fixed-capacity target pool and represents the mission state with a target-wise state matrix and a global state vector. A physics-constrained action mask constructs an admissible decision domain from current mission states, guiding policy learning toward physically executable decisions and avoiding infeasible exploration. An integrated training strategy facilitates policy optimization under coupled constraints. Comparative, ablation, and uncertainty experiments validate the effectiveness of PCM-PPO. PCM-PPO provides a feasible and interpretable scheme for autonomous mission planning of Titan deep space probes.

cs.AI · cs.RO

新论文 · v1arXiv:2611.15386

Actuation and Control of Passive Tendon-Driven Soft Extensions Through Rigid Manipulator Motion

Ignacio Cuiral-Zueco, Gonzalo López-Nicolás, Youcef Mezouar

原文摘要

Industrial manipulators offer reach and precision but can damage nearby structures in confined spaces. Soft robots provide compliant contact but typically have limited reach. Rigid-soft systems combine their advantages, although they often require dedicated actuation and control hardware. We propose actuating a passive tendon-driven soft extension through manipulator joint motion. As the robot moves, the exterior spans of tendons anchored to upstream links change, altering tendon tension and deforming the extension. Inspection through narrow apertures, for example, constrains the motion of the end effector (EE). We develop a controller that regulates the soft tip while preserving such EE constraints. A differential model separates rigid transport from soft deformation, combining analytic arm and tendon kinematics with a locally identified span-to-deformation map. We also derive a local upper bound on the number of independent deformation directions transmitted by the tendons under the EE constraint, and conditions for the exponential convergence of the tip error. A quadratic program with control barrier functions imposes obstacle clearance, fulcrum limits, and model-validity conditions. Simulations and real experiments show tip and shape regulation under free, position-constrained, and fulcrum-constrained EE motion.

cs.RO

版本更新 · v2arXiv:2610.06439

Better Call Reward: Reward Hacking as Strategic Abstention in Legal Reasoning Models

Subramanyam Sahoo, Justin Shenk

原文摘要

What happens when a legal AI model learns to look like a lawyer instead of reasoning like one? We fine tune Qwen3-8B with Group Relative Policy Optimisation (GRPO) against a proxy built from three surface features: citation count, legalese density, and response length. The model does not learn to reason more effectively. It stops committing to an answer. Across 16 yes or no legal reasoning tasks from LegalBench (N=320), overall accuracy collapses from 0.500 (chance) to 0.072 (McNemar p < 10^-36), driven entirely by the rate of properly formatted answers falling from 0.900 to 0.109. When the model does commit, accuracy is 0.657 against 0.556 before training, on only about 35 answers, suggesting the collapse is not mainly a loss of capability. The mechanism is visible in the training logs: the proxy keeps paying for length beyond what fits in the 256 token generation limit, completions grow until they fill that budget, and the required answer line is crowded out. We term this the Saul Goodman effect, a policy that becomes more lawyerly in form while becoming noncommittal, and prove formally that abstention is the optimal response to any surface feature proxy that attaches no penalty to it. A citation audit shows that structural plausibility checks cannot catch this failure: every case name we extracted from the trained model's held out completions passes our structural check, so citations must be verified against a case database. To detect this failure before deployment, we introduce three diagnostic tools: the Confidence Theater Score (CTS), the Citation Plausibility Rate (CPR), and the Regret Gap (RG), and report where each one falls short. In a domain where a confidently wrong answer can constitute malpractice, the broader lesson is direct: a reward function that measures how legal a response looks will produce a model that is maximally photogenic and minimally useful.

cs.LG · cs.AI · cs.CL · cs.CY