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 22:45 — 2026-11-16 22:45(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:2611.15325

CaseScope: A Multi-Field Benchmark for Case-Level Pathology Diagnosis with Multimodal LLMs

Wenchuan Zhang, Chengyang Zhang, Fengling Li, Ting Yin, Chen Shu, Jie Chen, Shuwan Zhang, Xiaoyan Wu 等 12 位作者

原文摘要

Pathologists rarely diagnose from a single image: they scan a section at low power for architecture, zoom into selected fields for cytology, and integrate the evidence across views and scales. Pathology benchmarks for multimodal large language models (MLLMs), however, mostly ask about one image or require whole-slide pipelines, and the few that show several fields draw nested regions from cancer slides of one public cohort. We introduce CaseScope, a benchmark of diagnosis from several curated fields of one lesion: CaseScope-Teaching, 436 cases from a residency-training image library, reconstructed, de-identified, harmonised against a WHO-based ontology and re-examined by a pathologist wherever an automated pre-review raised a doubt, and CaseScope-Atlas, 800 cases from 88 published atlases (387 of them figure series of one disease), labelled from the captions by two LLMs that must agree. They cover 843 entities across 18 organ systems and pose 4,679 items in six tasks, from open-ended and ten-way diagnosis with ontology-guided hard negatives under circular evaluation to the verification and location of morphological findings. Across twenty-eight open and four proprietary MLLMs, diagnoses are unstable -- the best open and proprietary models are correct under all ten option rotations on 17.9% and 61.7% of the teaching cases -- and wrong answers favour the classic look-alike; open models make little use of the additional fields on teaching cases, where the one-line specimen site matters more, although on atlas cases with three or more fields the fields count about as much as the site; and most open models verify findings barely above chance (median 62.0% vs. 50%, best 81.5%), gaining a median +3.4 points from images; proprietary models reach 83.2-93.2%. Rankings broadly replicate between the sources. The evaluation code is available at https://github.com/Wenchuan-Zhang/CaseScope.

cs.CV

版本更新 · v2arXiv:2606.10543

Flexible Flows for Biological Sequence Design

Yogesh Verma, Dani Korpela, Harri Lähdesmäki, Vikas Garg

原文摘要

Designing functional biological sequences requires navigating vast discrete spaces under strict evolutionary and biophysical constraints. Discrete Flow Matching (DFM) offers a generative framework over such spaces, but existing approaches rely on biologically uninformative couplings and offer limited flexibility for variable-length sequence generation and fine-grained control. We propose a structured coupling that encodes domain-specific preferences among sequence elements, biasing the source distribution toward plausible regions without modifying the flow objective or training procedure. Building on this, we introduce a latent edit-based rate parameterization that models variable-length generation via edit operations conditioned on a shared global latent, akin to a latent variable model, while remaining tractable. We further introduce a latent classifier-free guidance mechanism that steers generation coherently in continuous latent space, along with Dirichlet-prior temperature scaling for test-time control over edit operations. Our method achieves state-of-the-art performance across diverse biological sequence tasks, including density estimation, unconditional and conditional DNA sequence generation, and peptide sequence generation. Our code is available at https://github.com/Aalto-QuML/FlexFlow.

cs.LG · cs.AI · cs.ET · q-bio.QM

新论文 · v1arXiv:2611.15320

Delaunay-Based Nearest Neighbor Search with Grid-Accelerated Queries and Parallel Construction

Pengfei Wang, Shiqing Xin, Changhe Tu

原文摘要

Nearest-neighbor search is a fundamental primitive in computer graphics, with applications in distance field computation, surface reconstruction, and point cloud registration. In many graphics applications, data points lie on a surface manifold embedded in 3D space, whereas queries are distributed throughout the surrounding volume, a setting in which spatial partitioning structures such as k-d trees and grids can become inefficient. Dynamic-programming-based nearest-neighbor search (DP-NNS) offers an alternative: it traverses a successor DAG that records the history of incremental Delaunay construction and often outperforms spatial partitioning structures. However, its sequential construction is slow, and naive parallelization degrades query performance, limiting its practical adoption. We accelerate both queries and construction in DP-NNS. For queries, we introduce a grid-based initialization strategy that precomputes valid starting points on the successor DAG, shortening search paths while preserving exactness and enabling a twofold speedup. For construction, we analyze how parallel insertion affects query cost and develop a parallel algorithm that preserves logarithmic expected query cost under stated geometric assumptions. Our parallel construction algorithm achieves a 10x speedup over the sequential baseline while maintaining nearly the same query cost. We further reorder query points to improve cache locality. Extensive experiments demonstrate that our method outperforms widely used libraries in nearest-neighbor query performance.

cs.GR · cs.CG

新论文 · v1arXiv:2611.15317

BiPOP: Partial-Order Planning for Long-Horizon Bimanual Robotic Manipulation

Kun Chu, Xufeng Zhao, Jan-Gerrit Habekost, Cornelius Weber, Stefan Wermter

原文摘要

Bimanual robotic manipulation is versatile, yet its central challenge lies not only in executing dexterous motion with each hand, but also in coordinating the two arms for long-horizon manipulation under coupled spatial and temporal constraints. Prior work has largely focused on endowing robots with human-level manipulation skills, paying less attention to the structure of bimanual behaviour itself: many practical tasks are a coordinated composition of reusable unimanual skills and a small set of synchronous bimanual skills. This makes high-level spatio-temporal coordination, rather than skill acquisition alone, a key bottleneck for efficient bimanual execution. We propose BiPOP, a framework that couples Large Language Models (LLMs) with Temporal Partial-Order Planning for long-horizon bimanual manipulation. At its core is an explicit spatio-temporal model that partitions the workspace by arm reachability and captures inter-arm dependencies, arm assignment, synchronization requirements, and opportunities for safe concurrent execution. Building on this model, an LLM translates natural-language instructions into a formal PDDL specification, while a temporal planner enforces the constraints and schedules actions, yielding plans that are logically consistent at the symbolic level and efficient through parallel execution. On simulated long-horizon tasks of varying complexity, BiPOP outperforms open-loop planning with strong reasoning models and closed-loop ReAct-style planning in success rate, planning time, and plan efficiency, and the full pipeline runs end to end on the real bimanual robot NICOL. A failure analysis further shows that LLM-only planners mainly violate spatial and temporal coordination constraints, which motivates delegating these constraints to a symbolic planner. Code: https://github.com/Kchu/BiPOP

cs.RO

版本更新 · v2arXiv:2607.07922

Adversarial Decoys: Misdirecting Attention-Based Defenses in ViT

Giulia Marchiori Pietrosanti, Giulio Rossolini, Giorgio Buttazzo

原文摘要

Vision Transformers (ViTs) remain vulnerable to localized adversarial attacks, e.g., adversarial patches, while recent test-time defenses mitigate them by suppressing image tokens with abnormally high attention scores. These defenses exploit a strong coupling between attention and adversarial effectiveness: adversarial tokens often need to attract substantial attention to influence the prediction. We introduce adversarial decoys, image patches optimized to redirect attention, and therefore related defenses, towards selected target tokens. Rather than jointly optimizing misclassifications and defense evasion, our approach decouples the two objectives: the original adversarial region induces the incorrect prediction, while a separate decoy manipulates the attention ranking used by the defense. A layer-wise objective increases target-token attention and promotes these tokens above competing non-target ones. Since the decoy optimization is not tied to a specific attack formulation, the method can be easily integrated with any existing adversarial patch attack. Experiments on ImageNet across multiple ViT architectures and adversarial patch attacks show that decoys consistently redirect high attention scores away from the true adversarial region while preserving much of the attack effectiveness against attention-ranking defenses. These results reveal a fundamental limitation of using attention alone as an indicator of adversarial relevance.

cs.CV · cs.AI

新论文 · v1arXiv:2611.15314

AcrossPlace: Do Vision--Language Models Recognize Places Beyond Their Source Views? A Comprehensive Benchmark

Zhenyu Li, Tianyi Shang, Shuhan Dong, Ruotong Xu, Qingyu Zhu

原文摘要

Language-guided place recognition requires recovering geographic identity across visual observations beyond the image from which a description originates. However, aggregate retrieval scores often entangle observation transfer, gallery competition, and linguistic evidence, obscuring the underlying recognition capability. We introduce AcrossPlace, a provenance-aware benchmark establishing a systematic evaluation standard for bidirectional language--vision place recognition. Built upon GSV-Cities, AcrossPlace comprises 56,726 places across 23 cities, 156,800 automatically generated, rule-filtered descriptions, and 113,404 target images. Each description retains its source-image provenance, while retrieval operates on disjoint target observations. Text-to-image (T2I) and image-to-text (I2T) retrieval share a unified place-identity relevance criterion supporting multiple valid matches. Two geographic protocols evaluate unseen places within represented cities (P0) and generalization to held-out cities (P1). Paired observation comparisons, gallery-controlled evaluations, fixed-query scaling, and blinded annotation auditing systematically characterize retrieval performance. Across three training seeds, place-supervised SigLIP2 T2I Recall@1 decreases from 19.69% on source images to 12.31% on independent targets under P0. For fixed P1 queries, expanding the gallery from 500 to 8,425 places reduces Recall@1 from 14.67% to 2.87%. A 240-description audit further reveals discrepancies between automated acceptance and factual evidence. These findings establish AcrossPlace as a reproducible evaluation standard for assessing transferable place identity across visual observations and geographic environments.

cs.CV

版本更新 · v2arXiv:2611.09203

Copies Are Not Confirmations: How News Duplication Fools LLM Judges and Continual Learners

Abdelkader Metales, Miriam Grabenheinrich, Aicha Metales

原文摘要

A language model that keeps learning from the news sees the same claim many times, but many reports are copies of one source: wire stories, content shared within a media group, syndication and press releases. We ask whether LLM judges and continual learners treat such copies as independent confirmations. In 200,000 English articles from CC-News, a near-duplicate detector identifies copy links to another outlet for 18.9% of the articles, a detector-based lower bound. In a controlled experiment with claims about invented people, three identical reprints instead of one report raised the lies accepted by Llama 3.2 from 0.06 to 0.71 and by phi4 from 0.21 to 0.90; Gemini 2.5 Flash already believed a single report, and Gemini 2.5 Pro followed provenance information most consistently. In a simulated continual-learning stream, showing the judge one report per detected copy chain cut the copied lies accepted by 0.10-0.24 across four judges, at a mean cost of at most 4.2 points of accepted truths. Text similarity links none of the copies reworded by three LLMs. An entity-novelty test, the share of a report's named entities absent from an earlier look-alike, separates these rewrites from 100 author-labelled independent reports of the same event (AUC 0.98-0.99), statistically indistinguishable from TF-IDF and the best tested embedding model, but recognises only 17% of natural same-source derivatives. With both detectors run on 15,084 real articles carrying inserted claims, one report per chain cut the lies accepted by phi4 from 0.57-0.67 to 0.09 (copies), 0.14 (rewrites) and 0.34 (rewrites with invented details), and by Llama 3.2 from 0.56-0.62 to 0.07, 0.09 and 0.25, at a cost of 2.4 and 10 points of accepted truths. Counting reports or chains without a judge is unreliable, and colluding sources that do not copy each other remain an open problem.

cs.LG · cs.CL

新论文 · v1arXiv:2611.15311

Arbor: When Labels Are Not Coarsened at Random

Faisal Mohammad, Mohamed Younis

原文摘要

Merging datasets labelled at different levels of a taxonomy leaves some images with only a coarse label, and some classes with no fine-grained label at all. Standard training scores a coarse label by the total probability of the classes under it. Coarse labels name superclasses, and the standard objective is exact only when every class within a superclass is equally likely to receive a coarse label. We call a class novel if it is never labelled finely, and known otherwise. A novel class violates the assumption, and the classifier assigns its images to a known sibling. We propose Arbor, which models how labels become coarse. In each superclass with a novel class, Arbor adds a virtual novel class, and a logistic gate divides the superclass's probability between it and the known classes. The gate is trained with the likelihood of the observed labels under this model, the coarsening likelihood. We prove that at its population optimum the gate equals the Elkan-Noto correction of positive-unlabelled learning, and show that training the correction into the classifier improves on applying it after training. Across Cars, CUB and Aircraft with 90% coarse labels, Arbor correctly assigns 76-85% of novel-class images to their superclass, more than hierarchical classifiers, hierarchical novelty detection and the Elkan-Noto correction. Its top-1 accuracy over the leaf classes on known-class images is within 1.6 points of the best baseline, although its final predictions return the superclass more often. On the more challenging iNaturalist-2019 benchmark, Arbor improves novel-class accuracy from 46% (uniform target, the strongest baseline) to 68%.

cs.CV · cs.LG

新论文 · v1arXiv:2611.15309

Where Autonomy Belongs: A Separation of Powers for Agentic Program Evolution

Zhengyang Ai, Tong Liu, Jingxian Tang, Xiaobing Du, Tao Xiao, Fengshi Zhang, Zishuo Li, Xiaodong Ai 等 9 位作者

原文摘要

LLM-driven program evolution rests on three decisions: govern directs the search from accumulated evidence, propose constructs the next candidate, and judge settles validity and fitness. All three appear in every system, and existing work splits into two lines over who makes each. Controller-centric systems strengthen govern and neglect the autonomy of propose, which caps what evolution can reach. Agent-centric systems let a coding agent take over both govern and propose, at a cognitive and computational cost. Neither line protects the independent authority of judge, so benchmark reliability suffers. These failures share one root: autonomy is misallocated across the three decisions. We address this problem with EvoTrias, a separation-of-powers architecture in which each decision has one owner: Structured Governance for what spans evolution rounds, Open Proposal for construction inside a task, and Protected Judgment behind a private evaluator. These owners exchange typed contracts only, and runtime hooks reinforce the boundaries. This separation also lets EvoTrias run as a plugin that turns an existing coding agent into an evolution system. EvoTrias sets a new state of the art on mathematical, systems, and algorithm-design benchmarks. Three further results show what each owner buys: separating Governance from Proposal cuts token cost by 64%, Open Proposal extends evolution to a 680K-line industrial repository, and Protected Judgment keeps benchmark scores faithful to transfer.

cs.AI · cs.MA

新论文 · v1arXiv:2611.15307

Floating-point, exact and certified computation for control: paradigms, pitfalls, and how to combine them

Christina Katsamaki, Fabrice Rouillier

原文摘要

Many questions in linear systems theory have a discrete answer: is a polynomial Hurwitz, is a 2-D system structurally stable, is the peak gain finite, which solution of a Riccati equation is stabilizing? Algorithms implemented in floating point usually answer such questions, but often with a limited, or even no, guarantee on the result, which depends on a tolerance fixed a priori. Computer algebra and interval arithmetic provide other ways of computing, which give guarantees on the result, but usually at a higher cost. In this paper, we propose a short guide to these approaches: floating-point, exact, multiprecision and interval arithmetic, certified root isolation and certified Newton methods. Using small examples from control theory (each of them runs in a few seconds), we show what each approach guarantees and in which situations it fails, and how several approaches can be combined, for example by certifying a numerical solution, or by reducing a problem exactly before solving it with certified real solving. We then revisit the structural stability of 2-D systems (with and without parameters) and the computation of the $H_\infty$ norm, and we list the available implementations. The main examples are given as reproducible worksheets based on the PACE ecosystem.

cs.SC

新论文 · v1arXiv:2611.15306

Beyond the Explicit: Benchmarking and Detecting Implicit Harm in Short Videos

Chenyang Li, Qifan Yu, Saihui Hou, Zhaofeng He

原文摘要

Short videos have become a dominant online medium, where harmful content is increasingly conveyed in implicit and fragmented forms. Prior work largely focuses on explicit categories, such as violence and hate speech, and often fails to detect videos that subtly promote misleading values through narrative context, emotional framing, or cross-modal cues, partly due to the lack of dedicated benchmarks for implicit harm. To bridge this gap, we construct DeepHarm-7K, a large-scale harmful short-video dataset comprising 7,110 samples annotated according to a fine-grained harmful-content taxonomy. The dataset systematically covers implicitly harmful videos across diverse real-world scenarios and is constructed through a rigorous multi-stage quality-control process. Building on this dataset, we propose DeepHarm-VL, a multimodal detection framework that integrates visual, audio, and cross-modal reasoning. The framework employs a two-round reasoning strategy to progressively identify surface-level evidence and infer latent harmful implications. Experimental results show consistent improvements over strong baselines and state-of-the-art methods across multiple open-source video-language models, demonstrating its effectiveness in detecting both explicit and implicit harmful short videos.

cs.CV

版本更新 · v4arXiv:2504.16771

Projective Variety Recovery from Unknown Linear Projections

Yirmeyahu Kaminski

原文摘要

We study how a smooth irreducible variety $X$ of dimension $n$ and degree $d$ in $\mathbb{P}^m$, $m\geq n+2$, can be recovered from the equations of its images under two generic projections from unknown centers. This extends earlier results on curves in $\mathbb{P}^3$ and surfaces in $\mathbb{P}^4$ to any dimension. Since $X$ is the normalization of each image, one image determines $X$ modulo the projective group when $X$ is linearly normal; what two images add is the configuration, that is the pair of projections. For complete intersections we prove unconditionally that the set of configurations compatible with the two images and their Hilbert polynomial is finite, of size at most $|\mathrm{Lin}(X)|$, the projective automorphism group of $X$. The argument is effective and the bound is attained. We then introduce an equation generalizing Kruppa's equation, relating the duals of the images to a fundamental matrix, encoding the pair of projections modulo the projective group. We give a necessary condition for its solution variety to be finite, a bound on its isolated points, and a sufficient condition under which the true fundamental matrix is rigid: no quadric may contain a generic line $L$ together with the locus of points of $X$ whose tangent space meets $L$. Rigidity is unconditional for curves; in higher dimension it holds under a first-order liftability hypothesis, which we certify for a del Pezzo quartic surface in $\mathbb{P}^4$. Global finiteness of that solution variety remains open, necessarily so, as it contains solutions realized by no variety; for $\dim X<\mathrm{codim}\,X-1$ the obstruction amounts to knowing the fundamental matrix itself. Once the projections are known, $X$ is recovered from the two cones over the images: in codimension two their intersection is $X$ plus a component of degree $d(d-1)$; in higher codimension $X$ is its unique component of maximal dimension.

math.AG · cs.SC

新论文 · v1arXiv:2611.15300

Sharp Covering-Number Bounds for Feedforward ReLU Networks via Parameter Symmetries

Ruizhi Deng, Guohao Shen, Tangrui Xiong

原文摘要

In neural networks, distinct parameter configurations often realize the same input--output mapping, indicating that nominal parameter counts can overstate functional complexity. In this paper, we characterize the parameter-space symmetries of ReLU and Leaky ReLU networks across fully connected, convolutional, and residual architectures. We prove that every quasianalytic parameter map preserving the network function is a composition of positive channel scaling and channel permutation. By quotienting out both symmetries, we derive uniform covering-number upper bounds in which positive scaling reduces the complexity exponent from the total parameter count $S$ to $S-H$ (with $H$ denoting the number of scalable channels) and permutation symmetry divides the bound by the factorial count $G$ of channel permutations. We establish matching lower bounds showing that the exponent $S-H$ is sharp in general and that the divisor $G$ is attained for one-hidden-layer and generically identifiable networks. We further extend these results to general continuous piecewise-linear activated networks. Finally, we apply our covering-number bounds to derive statistical risk guarantees for empirical risk minimization and capacity control beyond i.i.d.\ settings, and show how the underlying symmetries help explain weight rebalancing in function approximation and flat directions in empirical risk optimization.

cs.LG · math.ST · stat.ML

新论文 · v1arXiv:2611.15299

Small classes have linear neighbourhood complexity

John Sylvester, Viktor Zamaraev, Maksim Zhukovskii

原文摘要

A hereditary graph class is small if it contains at most $n!c^n$ labelled graphs on $n$ vertices for some constant $c$. Small classes include proper minor-closed classes and classes of bounded twin-width. We prove that every small class has linear neighbourhood complexity, that is, in every graph in the class, vertex neighbourhoods intersect any $k$-vertex set in only $O(k)$ distinct ways. This resolves a conjecture of Bonnet, Duron, and the first two authors (ITCS 2025) and makes the algorithmic, structural, and representation consequences of linear neighbourhood complexity available for every small class. In particular, every small class admits an adjacency labelling scheme with $O(\log^2 n)$-bit labels, improving the previous $O(\log^3 n)$ bound.

math.CO · cs.DM

新论文 · v1arXiv:2611.15298

Quantum Recurrent Architectures for Financial Time Series Prediction

Danyal Maheshwari, Gerhard Hellstern, Martin Zaefferer, Martin Braun, Tanja Döhler

原文摘要

This study evaluates quantum classical hybrid recurrent architectures for financial time-series forecasting: a Quantum Long Short-Term Memory (QLSTM) network and Quantum Reservoir Computing (QRC) with lag-based inputs. Classical lag vectors are encoded into quantum states via amplitude encoding, which compactly represents normalized observations but requires nontrivial state preparation; the QLSTM gates and the QRC reservoir are implemented with parameterized quantum circuits, while classical optimizers train the interface and readout layers. On a monthly revenue series extended to 1056 months by a Gaussian-process--hidden-Markov generator, with a fixed chronological 70/15/15 split and one-step-ahead forecasts, we compare both models against a full-size LSTM, a classical control matched in parameter budget and interface architecture, a small plain LSTM, and classical reservoir computing, over ten seeds and against persistence, seasonal-naive and SARIMA benchmark forecasts. The QLSTM, with 2,105 parameters, is statistically indistinguishable from an LSTM 126 times larger, but equally indistinguishable from the matched classical control and from a plain LSTM of the same budget, so its parameter efficiency cannot be attributed to the quantum circuits. The quantum reservoir is worse than a classical reservoir of equal feature dimension in all ten seeds and worse than SARIMA, whereas all recurrent models and both classical reservoirs improve on SARIMA by about 2\% in RMSE. In a 24-month rolling-origin backtest the classical LSTMs degrade more slowly than the QLSTM. We find no evidence of a quantum accuracy or parameter-efficiency advantage in this low-dimensional setting.

quant-ph · cs.AI

新论文 · v1arXiv:2611.15296

APG-Planner: Learning Vision-Based High-Speed Trajectory Planning via Analytic Policy Gradient

Jinwoo Lee, Geunsik Bae, Changseung Kim, Jiwoo Kim, Woojae Shin, Hyondong Oh

原文摘要

Vision-based high-speed navigation requires a trajectory planner that can reactively generate smooth, collision-free trajectories from onboard visual observations. Deep reinforcement learning offers a promising approach to vision-based trajectory planning, as it enables the optimization of sequential decision-making policies in complex environments. However, model-free policy gradient methods rely on sample-based policy gradient estimation, which yields noisy gradient estimates and inefficient training. To this end, we propose a framework for learning trajectory planning policies using the analytic policy gradient (APG) method and use it to train APG-Planner, a vision-based high-speed trajectory planner. The proposed framework leverages differential flatness to analytically reconstruct vehicle states from planned trajectories, thereby enabling differentiable trajectory-level state transitions without requiring an explicit low-level dynamics model. By iteratively performing trajectory planning and differentiable state transitions, we construct a differentiable episode rollout and analytically compute policy gradients from episode-level losses, improving both training efficiency and policy performance. After approximately 30 minutes of training on a single GPU, APG-Planner demonstrates comparable or superior performance to existing baseline methods in simulation and achieves speeds of up to 11\,m/s in real-world flight experiments. We further validate the proposed framework on a skid-steer ground robot, demonstrating its applicability to differentially flat robotic systems beyond quadrotors.

cs.RO

新论文 · v1arXiv:2611.15291

Learning Rate Transfer and Feature Learning Across Depth for Constrained Spectral Optimizers: Complete Scion

Eline Pot, Laurent Lefebvre, Mathieu Segond, Antonio Silveti-Falls

原文摘要

Hyperparameter transfer across model depth is instrumental in training deep models, as it enables learning rates calibrated on shallow networks to be effectively applied to deeper architectures. The recent CompleteP framework establishes a $1/L$ residual scaling rule that demonstrates this for Adam-family optimizers. We study whether similar transfer extends to constrained optimization methods, specifically Frank-Wolfe and projected non-Euclidean normalized steepest descent, where updates are determined by linear minimization oracles. We derive scaling rules for these methods which confirm that the $1/L$ residual scaling applies, although the arguments differ from the Adam-family optimizers. Our analysis supports depth-wise transfer and width-wise transfer of the learning rate for the Scion algorithm while ensuring feature learning and avoiding the kernel-regime. In experiments on NLP transformers (nanoGPT on FineWeb), vision transformers (DeiT on ImageNet), and vision MLPs (MLP-Mixer on CIFAR-10), we observe learning rate transfer across depth and feature learning for appropriately chosen residual scaling.

cs.LG · math.OC

版本更新 · v2arXiv:2605.15729

Constructing Generalizable Parallel Algorithm Portfolios for Multi-Objective Binary Optimization via General-Purpose Instance Generation

Zhiyuan Wang, Shengcai Liu, Shaofeng Zhang, Ke Tang

原文摘要

Despite progress in constructing generalizable parallel algorithm portfolios~(PAPs), no general-purpose method is yet available for multi-objective binary optimization problems~(MOBOPs) without problem-specific customization. To fill this gap, this work proposes domain-agnostic co-evolution of parameterized search for multi-objective binary optimization~(DACMO), the first general-purpose method for automatic PAP construction on MOBOPs. Its core, Unified Instance Representation~(UIR), represents and generates MOBOP instances with varying dimensions without knowledge of problem-specific instance structures. UIR separates common features of the target class from instance-specific features, preserving the common features learned from real instances during generation. DACMO co-evolves the training instances and PAP by alternating UIR-based instance augmentation with PAP updates. Experiments cover four structurally different MOBOP classes, including two black-box classes from real-world applications. Starting from five real training instances per class, DACMO constructs high-performance PAPs without problem-specific customization. On unseen instances from all four classes, these PAPs outperform portfolios of tuned classical multi-objective evolutionary algorithms and those constructed by the recent automatic PAP construction method EoH-S. DACMO also achieves overall performance comparable to the strong CEPS baseline and significantly outperforms it on two classes, without the substantial expert effort CEPS requires for problem-specific instance generator design.

cs.NE

新论文 · v1arXiv:2611.15288

FineSpeaker: Reliable Speaker Representation Learning at Fine Temporal Resolution

Yongkang Yin, Wenhao Yang, Lichun Fan, Jian Luan, Hang Su, Yuexian Zou

原文摘要

Speaker representation learning is fundamental to speech processing tasks such as speaker verification and diarization. Although long-context speaker encoders can produce stable and discriminative embeddings, short-duration verification and diarization require finer-grained temporal representations. Short temporal contexts limit the available speaker evidence and make speaker identity difficult to disentangle from phonetic and acoustic variability. This creates a temporal resolution-representation reliability trade-off: speaker representations should maintain local within-speaker consistency, between-speaker discriminability and remain sensitive to speaker changes at fine temporal resolution. Accordingly, we propose a framework for reliable speaker representation learning at fine temporal resolution. FineSpeaker first constructs resolution-controllable frame-wise speaker representations through mean pooling over an adjustable n-frame window, enabling flexible control of representation granularity. It then introduces a speaker encoder with local temporal modeling that exploits short-range speaker context to improve the stability and discriminability of short-duration and frame-wise representations. Furthermore, we develop a generative disentanglement strategy for optimizing frame-level speaker representations, combining VAE-based probabilistic modeling with flow matching to characterize speaker-related variation and suppress non-speaker interference. Experiments demonstrate competitive full-utterance performance, improved fine-grained speaker verification on VoxCeleb and CN-Celeb, and strong zero-shot diarization results, highlighting the benefits of balancing temporal resolution and representation reliability.

cs.SD

版本更新 · v4arXiv:2605.26078

Global Convergence of Wasserstein Policy Gradient for Entropy-Regularized Reinforcement Learning

Zhaoyu Zhu, Rui Gao, Shuang Li

原文摘要

Wasserstein policy gradient (WPG) optimizes stochastic policies in the optimal-transport geometry of action distributions. In entropy-regularized reinforcement learning, WPG combines transport driven by gradients of the soft Q-function with entropy-induced diffusion. The resulting dynamics are coupled through Bellman policy evaluation, and the policy optimization objective is generally nonconvex. We establish a unified global convergence theory of WPG for infinite-horizon discounted Markov decision processes with general state spaces and continuous action spaces. We identify a statewise gradient-dominance inequality by exploiting the Bellman structure together with log-Sobolev inequalities, which we show to hold uniformly along the trajectory. This inequality relates the squared norm of the Wasserstein policy gradient to the value gap in the sup norm. In continuous time, we prove well-posedness of the flow and exponential convergence of the value gap in the sup norm and of the policy in statewise KL divergence. In discrete time, by additionally controlling the regularity and the discretization error, we obtain geometric convergence to a neighborhood of the optimum with constant step sizes and an $O(\varepsilon^{-1})$ iteration complexity with diminishing step sizes for fixed problem parameters. Finally, we establish faster local convergence rates for both dynamics, governed by the geometry of the optimal policy.

cs.LG

版本更新 · v2arXiv:2611.12305

HumanSightBench: A Benchmark for Human Visual Perception in MLLMs

Zixin Su, Yangguang Lin, Zegang Peng, Yuntong Sun, Xuanyi Wang, Jiahao Wang, Qiankai Xu, Bing Yuan 等 13 位作者

原文摘要

Multimodal large language models (MLLMs) have demonstrated strong performance on knowledge-intensive and reasoning-based vision-language benchmark datasets, but still struggle to solve the visual questions that people can solve easily. Evaluating the differences between MLLMs and human-level visual perception is challenging due to end-to-end evaluation which combines various skills, or because we are uncertain if individual items are valid. Specifically, observed errors may reflect visual perception, language priors, background knowledge, question ambiguity, and underspecified answers. Our guideline for HumanSightBench is that we construct items that are human-easy by construction and model-hard by evaluation. HumanSightBench contains 2,876 visual questions with short answers across 7 top-level perceptual categories and 21 subcategories. We only include items that pass five inclusion gates: question clarity, answer uniqueness, visual grounding, human solvability, and text-only shortcut filtering. According to the protocol, errors are defined as incorrect answers on human-validated visual tasks with the evaluated interface. The inclusion gates are intended to reduce confounds from ambiguity and external knowledge. The best pass@3 score is 65.6% across eight evaluated MLLMs. No evaluated model exceeds 27.5% on Counting. These results identify capability gaps that are hidden by overall scores and can help future training experiments and held-out evaluation.

cs.CV · cs.AI · cs.CL

新论文 · v1arXiv:2611.15284

A hybrid machine learning-analog framework for downscaling unprecedented climate extremes

Nina Effenberger, Florian E. Roemer, Maybritt Schillinger, Maxim Samarin, Erich Fischer, Reto Knutti

原文摘要

Extreme weather events are critical for climate impact assessments. While physical models can simulate unprecedented extremes, for example, through ensemble boosting, this approach is computationally demanding, often restricting output to coarse spatial resolutions. Probabilistic generative downscaling offers a solution by enabling efficient sampling from the entire climate distribution, including its extreme tails. Here, we first demonstrate that machine learning generated regional climate model heatwaves over Paris are physically realistic. We then leverage an emulated machine learning dataset to perform analog-based downscaling of coarse global climate model ensemble boosting events. Our hybrid framework effectively bridges high-resolution regional detail with rare-event sampling, providing an efficient tool for extreme event downscaling and capturing multi-model uncertainty.

cs.LG · stat.AP

版本更新 · v2arXiv:2602.02505

Learning-augmented smooth integer programs with PAC-learnable oracles

Hao-Yuan He, Jiang-Tian Xue, Ming Li, Zhi-Hua Zhou

原文摘要

We study an end-to-end learning-augmented framework for smooth integer programs, a broad class of Boolean polynomial optimization problems including MAX-CUT and MAX-k-SAT. An oracle learned from sampled instances predicts a full Boolean assignment for each unseen instance. For fixed-degree objectives, our polynomial-time pipeline, LEAP, linearizes the objective around the predicted assignment, solves an error-budgeted linear program, and rounds the fractional solution. We establish two guarantees. First, for any prediction, the approximation loss of LEAP degrades gracefully with prediction error: it is controlled by an error footprint that quantifies how much mispredicted coordinates can perturb the objective's local coefficients. Second, we bound the pseudo-dimension of the LEAP-induced cost class, yielding a PAC-style guarantee for learning an oracle from a predictor class via empirical risk minimization on instances from an unknown distribution. Together, these results give an end-to-end bound that separates the expected footprint loss of the best oracle in the class from the statistical error of learning. In experiments on MAX-CUT and MAX-SAT, LEAP substantially repairs imperfect predictions, empirical risk minimization selects effective oracle candidates, and the pipeline scales to application benchmarks with up to two million clauses.

cs.DS · cs.AI · cs.LG

新论文 · v1arXiv:2611.15281

Priority-Based Network Access with Strict-Priority Scheduling in PLCA Multidrop Networks

Mohamed Elattar, Zhe Lou, Georg Carle

原文摘要

Shared-medium Ethernet is gaining renewed attention across industrial domains, driven by the adoption of 10BASE-T1S. In such multidrop networks, collision-free and deterministic medium access is achieved through Physical Layer Collision Avoidance (PLCA). However, PLCA follows a node-based cyclic access structure that does not account for traffic priority, leading to high access latency for time-sensitive traffic and limited scalability with increasing network size. Existing PLCA extensions improve latency for selected devices but remain restricted to node-based access and do not support network-wide traffic differentiation. To address these limitations, this paper introduces a priority-based network access and scheduling mechanism for multidrop networks. The proposed approach decouples medium access from node-based ordering and enables class-based prioritization, allowing high-priority traffic from any device to be transmitted ahead of lower-priority traffic. It further supports a configuration that emulates Strict Priority (SP) scheduling, providing low and bounded access latency for highest-priority traffic. Analytical bounds on the worst-case access latency are derived, validated, and shown to be tight. Performance evaluation demonstrates that the proposed mechanism reduces the maximum observed access latency for high-priority traffic by over 65% compared to PLCA, with latency increasing by less than 1% as the network grows from 8 to 40 devices.

cs.NI

新论文 · v1arXiv:2611.15280

Autonomous Trail Racing with a Bicycle Robot

Fabian Jenelten, Deepana Ishtaweera, Pablo Garcia, Liana Margolese, Igor Bogoslavskyi, Luca Buonocore, Jan Migliazza, Aleksandrs Ecins 等 13 位作者

原文摘要

Navigating off-road trails on a mountain bike requires diverse locomotion skills, including aerial maneuvers, cornering, and slip recovery, paired with sophisticated perception for scene understanding. Translating these capabilities to autonomous, two-wheeled robots remains a formidable challenge. We show that autonomous, high-speed trail navigation can be achieved through a single idea: encoding the trail as a semantic layer of the terrain. In simulation, this layer embeds the navigation task into the training environment; during deployment, it is reconstructed onboard from camera and LiDAR data. First, we introduce a unified learning framework that treats local navigation and locomotion as a joint optimization problem, bypassing the twist or pose command interface of classical control hierarchies. With fixed reward weights, the same framework produces both trail racing and agile parkour, shaped only through terrain semantics and episode length. Second, our perception pipeline combines a frozen vision foundation model with a lightweight dual-head decoder that predicts per-pixel terrain semantics and confidence scores, which are fused with LiDAR range measurements into an egocentric semantic elevation map at 30 Hz onboard an NVIDIA AGX Orin. We validate our framework on the Ultra Mobility Vehicle (UMV), a bicycle robot that autonomously raced across all sections of a kilometer-long, unstructured mountain bike trail at speeds exceeding 5 m/s and averaging 3.26 m/s. In simulation, our policy fails up to an order of magnitude less often than hierarchical baselines while driving faster.

cs.RO

版本更新 · v3arXiv:2510.20454

Capturing Intransitive Dominance in Tennis Forecasting: A Graph Neural Network Approach

Lawrence Clegg, John Cartlidge

原文摘要

Intransitive player dominance, where player A beats B, B beats C, but C beats A, is common in competitive tennis. Yet, there are few known attempts to incorporate it within forecasting methods. We address this problem with a graph neural network approach that explicitly models these intransitive relationships through temporal directed graphs, with players as nodes and their historical match outcomes as directed edges. Our model (65.7% accuracy, 0.214 Brier score) forecasts competitively with established rating systems such as Weighted Elo. Although it does not improve on the baseline in unconditional accuracy, a forecast-encompassing test shows that it carries complementary information. A combined forecast significantly outperforms Weighted Elo, and there is some indication that the gain is larger on the intransitive matchups our model targets. A graph-based representation of player interactions thus captures a forecasting signal that transitive rating systems discard, even between players who share no common opponents.

cs.LG

新论文 · v1arXiv:2611.15277

DroneDiffusion: Uncertainty-Guided Diffusion Path Planner for UAV Navigation

Ivan Valuev, Iana Zhura, Valerii Serpiva, Miguel Altamirano Cabrera, Dzmitry Tsetserukou

原文摘要

We present DroneDiffusion, a perception-aware diffusion planner for local 3-D UAV trajectory generation that explicitly models and avoids failure zones caused by lens flare, overexposure, and reflective surfaces. A vision-language perceptual safety estimator produces a pixel-aligned reliability map that highlights regions of degraded sensing. A denoising diffusion probabilistic model (DDPM) is conditioned on this map together with the onboard RGB image, goal point, and initial velocity to generate egocentric 3-D trajectories. To enforce obstacle avoidance at inference time without retraining, we apply training-free CHOMP guidance over a 3-D Euclidean signed distance field. The model is trained entirely in simulation and deployed zero-shot on a real drone. Reliability conditioning raises the mean perceptual reliability along generated trajectories to 0.71, compared with 0.63 for a geometric baseline, while guidance halves the collision rate of the learned prior from 14.7% to 7.8%. Perceptual safety conditioning and geometric guidance are therefore complementary: treating sensor reliability as a first-class signal reduces incursions into unreliable regions while maintaining competitive navigation performance.

cs.RO

新论文 · v1arXiv:2611.15276

Verify by Doing: Co-Evolving Agent Execution and Evaluation without External Feedback

Yitong Lin, Jianda Chen, Yihao Hu, Pan Wang, Hang Wang, Zhihao Wen, Xiaoxue Ren, Wei Wu

原文摘要

Agent self-improvement requires feedback about task success, yet external feedback may be unavailable at test time. Agents must then infer success from observable evidence, and incomplete verification can misguide subsequent updates. We introduce Verify by Doing (VbD), a method that co-evolves Actor and Evaluator harnesses within a single task while keeping model weights fixed. The Evaluator actively inspects artifacts and environment state with tools; its findings ground an internal signal for output selection and guide updates to both execution and verification procedures. Proposed Evaluators undergo comparative review on common candidates, while Actor updates require support from their own fresh outputs. Across AutomationBench, $τ^3$ Banking, and selected GDPval tasks, VbD achieves the highest overall scores among the compared methods operating without external feedback, with gains of 3.3, 4.1, and 2.3 percentage points over the strongest baselines on the main backbone. Ablations support the contributions of active verification and Evaluator evolution, and selection from fixed candidate pools shows that improved verification can recover successful outputs without generating new ones.

cs.AI

版本更新 · v2arXiv:2606.16175

PAL-Bench: Evidence-Grounded Profile Reconstruction from Longitudinal Personal Albums

Qiwei Yan, Zhiqiang Yuan, Zexi Jia, Nanxing Hu, Kailin Lyu, Jie Zhou, Jinchao Zhang

原文摘要

Longitudinal personal albums are weak-schema multimodal databases: noisy perceptual records whose key facts require joins across faces, text, timestamps, locations, and repeated events. Existing visual, video, document, and lifelog benchmarks test sub-problems, but not album-scale profile reconstruction with social identity binding and evidence citation. Benchmarking this task is difficult because the ground truth needed for evaluation--owner profiles, social graphs, face-name maps, and evidence provenance--is private state that real albums cannot safely release. We introduce PAL-Bench, a controlled benchmark for evidence-grounded reconstruction under a public-record contract. Its Evidence Compiler builds latent private worlds, programs target-level evidence paths, renders album pixels, re-measures them through perception pipelines, and exports audited public/private views. Agents receive only perception-derived public records; targets, identifier maps, and evidence paths remain hidden. PAL-Bench contains 50 synthetic users, 36,659 public photo records, and 2,799 targets over owner facts, identities, and relations. A privacy-preserving audit with 10 participants confirms that PAL-Bench evidence structures match real private albums, though equivalent releases remain privacy-prohibitive. Across seven systems and two compute-matched diagnostics, a seven-metric protocol reveals a gap between plausible profile summarization and faithful social reconstruction: systems recover some owner facts but struggle with recurring identities and evidence citation. PAL-TRACE, a reference framework that freezes identity bindings before owner-fact mining, performs best but leaves hard identity resolution far from solved. PAL-Bench provides a testbed for perceptual entity resolution, multimodal data integration, temporal evidence aggregation, and provenance-aware structured prediction.

cs.AI

版本更新 · v2arXiv:2611.14498

Measuring Robustness and Efficiency in a Connectome-Constrained Fly Visual System Model: Linking an Adversarial and a Perceptual Illusion

Alexandre Garcia-Duran, Alejandro Rodriguez-Garcia, Manuel Molano-Mazón, Alexandre Hyafil, Srikanth Ramaswamy, Janne K. Lappalainen

原文摘要

Biological visual systems compute reliably under tight metabolic and wiring constraints, whereas artificial vision systems consume far more energy and can be sensitive to adversarial perturbations. Whether networks with brain-like wiring are more robust at lower computational cost is unclear. We address this by benchmarking a task-optimized, continuous-time recurrent neural network of the fly visual system constrained by its synaptic connectome (FlyNet) against artificial neural networks. We compare it against CNNs, RAFT, and continuous-time recurrent neural networks without detailed connectome constraints (sCTRNN, CTRNN) on a collision-detection task within a differentiable 3D rendering pipeline. FlyNet is adversarially vulnerable but, relative to its compute, more robust to common video corruptions than RAFT, while using up to five orders of magnitude fewer FLOPs. The CTRNNs remain more robust than FlyNet across almost all tested perturbations, suggesting that connectome constraints benefit efficiency, through their low edge count, over robustness. Recurrence improves robustness to white-box attacks in RAFT and temporal context improves robustness to common corruptions. FlyNet's position and velocity representations degrade under adversarial textures but recover under counterfactual textures that restore correct classification while staying equally far from training, ruling out unfamiliar textures as the cause of failure. The reverse-phi illusion similarly disrupts these representations and inverts the same low-level motion statistic as the adversarial textures. Optimizing this input statistic alone, without access to the network, suffices to induce classification failure, attributing both perturbations to the same circuit vulnerability. Connectome-constrained models can thus surface adversarial failure mechanisms as testable hypotheses about biological perceptual vulnerabilities.

cs.CV · q-bio.NC