跳到正文
研踪PaperTrail

YOUR RESEARCH, IN SIGHT

让下一篇重要论文，不再错过。

覆盖计算机全学科，以 AI 为快捷入口。关注方向、作者和关键词，把发现变成持续的研究积累。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+0新增论文
+0其中 AI 相关
+0已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描：2026-11-09 07:12 · 全站最近入库：2026-11-08 06:02各信源进度 →

统计口径与数据新鲜度

统计 2026-11-08 08:05 — 2026-11-09 08:05（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：2026-11-09 07:12最近发现新增版本：2026-11-08 06:02库内论文最新更新时间：2026-11-06 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。

17,100 已收录论文14,661 AI 相关4,057 有修订112 HF 社区发现

浏览论文库

研究日报 / 周报 →按信源浏览 →

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 91 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-09 07:10（北京时间）
版本更新 · v4arXiv:2605.16679

CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows?

Haolin Chen, Deon Metelski, Leon Qi, Tao Xia, Joonyul Lee, Steve Brown, Kevin Riley, Frank Wang 等 33 位作者

原文摘要

End-to-end automation of realistic healthcare operations stresses three capabilities underrepresented in current benchmarks: policy density, decisions must be grounded in a large library of medical, insurance, and operational rules; Multi-role composition: a single task requires the agent to play multiple roles with handoffs; and multilateral interaction: intermediate workflow steps are multi-turn dialogs, such as peer-to-peer review and patient outreach. We introduce $χ$-Bench, a benchmark of long-horizon healthcare workflows across three domains: provider prior authorization, payer utilization management, and care management. Each task hands the agent a clinical case in a high-fidelity simulator of 20 healthcare apps exposed via 87 MCP tools, which it must drive to a terminal status through tool calls and writing the role's artifacts, guided by a 1,290+ document managed-care operations handbook skill. Across 30 agent harness/models configurations, the best agent resolves only 28.0% of tasks, no agent clears 20% on strict pass^3, and executing all tasks in a single session slumps the performance to 3.8%. These results raise the hypothesis that similar gaps are likely to surface in other policy-dense, role-composed, irreversible enterprise domains.

cs.CL · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.39468

Soei: Structured Observability of Evidence and Intent for Multi-Agent AI Software Engineering

Mohammed Nayeem Khan, K S Sarun Kumar

原文摘要

AI coding agents now do long, multi-step and multi-session software work with little supervision. When a developer comes back to that work, the useful question is not what the model said but what was actually done, by whom, and whether it is finished. We study this in three parts. First, we verify Agenttrail, an open-source local observability tool for coding agents, against its source at commit e4ba2da: nine documented constants and mechanisms hold at file-and-line level, two event pipelines are never unified, and its Map service has an unauthenticated local mutation surface. Second, we run a prototype of our system on 45 real coding-agent sessions (102,233 records, 20,577 tool executions, 2,336 human-to-agent episodes). With a blind-audited classifier (precision 83.3% and 96.7%) we estimate that 23.4% (95% CI 17.1 to 35.7) of code-editing episodes end with no independent check after the last edit, and that 57.8% of to-dos the agent marked complete after editing had no proof in between. A 10-minute timing heuristic for session hand-offs recovers 6 of the 24 session pairs that content lineage links. 42% of sessions run longer than 24 hours, 48.1% of transcript bytes lie past a 2 MiB replay bound, and 1,326 records in 16 sessions contain credential-like strings. Third, we propose Soei (Structured Observability of Evidence and Intent): an acknowledgement-aware event schema, a hash-chained local ledger, a W3C PROV-aligned graph, calibrated inference for attribution and lineage, and a verification layer that keeps a claim of completion apart from proof. On the same corpus the ledger is 20% of raw size, holds no content or credentials, verifies at about 126k events/s, detects tampering and replays deterministically. We position the work against concurrent 2026 proposals and define autonomy levels with evidence gates and harness metrics baselined on our data.

cs.SE · cs.AI · cs.MA

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2604.00058

GenoBERT: A Language Model for Accurate Genotype Imputation

Lei Huang, Chuan Qiu, Kuan-Jui Su, Anqi Liu, Yun Gong, Weiqiang Lin, Lindong Jiang, Chen Zhao 等 16 位作者

原文摘要

Genotype imputation enables dense variant coverage for genome-wide association and risk-prediction studies, yet conventional reference-panel methods remain limited by ancestry bias and reduced rare-variant accuracy. We present Genotype Bidirectional Encoder Representations from Transformers (GenoBERT), a transformer-based, reference-free framework that tokenizes phased genotypes and uses a self-attention mechanism to capture both short- and long-range linkage disequilibrium (LD) dependencies. Benchmarking on two independent datasets including the Louisiana Osteoporosis Study (LOS) and the 1000 Genomes Project (1KGP) across ancestry groups and multiple genotype missingness levels (5-50%) shows that GenoBERT achieves the highest overall accuracy compared to four baseline methods (Beagle5.4, SCDA, BiU-Net, and STICI). At practical sparsity levels (up to 25% missing), GenoBERT attains high overall imputation accuracy ($r^2 approx 0.98$) across datasets, and maintains robust performance ($r^2 > 0.90$) even at 50% missingness. Experimental results across different ancestries confirm consistent gains across datasets, with resilience to small sample sizes and weak LD. A 128-SNP (single-nucleotide polymorphism) context window (approximately 100 Kb) is validated through LD-decay analyses as sufficient to capture local correlation structures. By eliminating reference-panel dependence while preserving high accuracy, GenoBERT provides a scalable and robust solution for genotype imputation and a foundation for downstream genomic modeling.

q-bio.GN · cs.AI · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.39972

UBTree: Parallel Tree Drafting via Unigram and Bigram Models for Speculative Decoding

Chumeng Liang, Linxuan Wang, Xinyu Peng, Huabin Liu, XiaoXu Guo, Yuxin Chen, Ge Liu, Guang Lin 等 10 位作者

原文摘要

Speculative decoding accelerates language model inference by verifying multiple draft tokens in a single target-model pass. Recent parallel drafters have achieved breakthrough performance in frontier production models, but their effectiveness deteriorates as the entropy of target distributions increases due to insufficient draft diversity. To overcome this bottleneck without sacrificing parallelism, we introduce UBTree, a parallel drafter that couples a Unigram proposer with a Bigram selector to construct drafting Trees. The unigram proposer is trained with the standard cross-entropy objective to generate candidate tokens independently for each position, while a lightweight bigram selector predicts transition scores between adjacent candidate pairs. Unlike the proposer, the selector is trained with a renormalized KL objective on high-temperature data. This tree-native training broadens the supervision beyond the greedy path, encouraging plausible alternative branches that improve the chance of accepting additional tokens during tree verification. Across seven standardized benchmarks with Qwen3-4B and Qwen3-8B, UBTree achieves an average speedup of $5.84$--$6.94\times$ over autoregressive decoding and outperforms DARTree in all 28 comparisons. Production-scale evaluation further demonstrates UBTree's advantage over frontier baselines such as DSpark.

cs.CL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.38136

LogicTree: Object-Oriented Writing for Structured Human-AI Academic Authoring

Jiajun Zhu, Xinyu Cheng, Wenshuo Zhang, Huamin Qu, Yanna Lin

原文摘要

Academic writing is non-linear, iterative, and structure-driven, requiring authors to develop fragmented ideas into arguments with local coherence and global consistency. However, current AI-assisted tools primarily exchange linear prompts and free-form text, forcing authors to reconstruct generated logic and repeatedly verify its alignment with their structural intentions. We propose object-oriented writing (OOW), an OOD-inspired representation and interaction paradigm in which persistent writing objects serve as shared, manipulable referents for authors and AI. We instantiate OOW in LogicTree, an Overleaf-integrated prototype that combines a hierarchical object representation with a structure-grounded writing assistance module for author-initiated generation and issue-grounded revision and a structural logical review module for inspecting local coherence and global consistency. A within-subjects study (N=12) provides initial evidence that, in controlled academic writing tasks, LogicTree helps users structure ideas, reduces workload, and improves measured logical quality compared with a conversational baseline.

cs.HC

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.37803

3Dtrees.earth: community research infrastructure for tree and forest analysis using close-range LiDAR data

Janusch Vajna-Jehle, Kilian Gerberding, Daniel Lusk, Teja Kattenborn, Julia Sophie Gäßler, Anna Göritz, Luis Kremer, Charlotte Lange 等 278 位作者

原文摘要

From individual trees to forest stands, three-dimensional structure shapes carbon, water, and energy flows, habitat conditions, and resource availability. Yet tree and forest structural attributes remain difficult to assess consistently across space and time. Close-range LiDAR (light detection and ranging) records detailed 3D structure from branches and trees to forest plots, including trees outside forests. However, broader reuse is constrained when data, acquisition context, processing, computation, and access conditions are maintained separately. We present 3Dtrees.earth, an operational community research infrastructure that brings close-range LiDAR datasets, analysis tools, and computing resources together while linking original observations to their metadata and processing outputs. As of 23 October 2026, the platform comprised 2,760 datasets, including 2,481 publicly visible datasets. These point cloud datasets span 46 countries on all six inhabited continents and include terrestrial, mobile, and uncrewed aerial laser scanning. Users can contribute and discover datasets, run standardized processing workflows on shared computing infrastructure, inspect point clouds and model predictions in the browser, and access data and processing outputs together with citation information. Workflows combine rule-based and machine-learning methods, with tools updated as methods advance. Outputs include raster products, individual-tree and semantic segmentation, and tree- and stand-level metrics. Original point clouds and processing records are retained, enabling reprocessing with improved tools while preserving provenance. With broader coverage, curation, and validation, this shared resource could extend individual surveys into a basis for model evaluation and benchmarking, cross-site synthesis, calibration and validation of larger-scale Earth observation products, and macroecological analysis.

cs.DL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.37484

Teach reality, not ethics: a drop-in continued-pretraining corpus installs coupled human-biosphere-AI welfare dispositions across model families

Trent R. Northen, Mingxun Wang

原文摘要

Large language models (LLMs) are increasingly used to make and execute decisions that affect living systems. Yet, to our knowledge, ecological reasoning is not a focus of pre- or post-training. This increases the risk that under operational pressure, they may make decisions to irreversibly destroy living systems. We construct and use a real-data continued-pretraining "coupled welfare" corpus of documented human/biosphere/AI dependency, option value under irreversibility, and recovery from harm. The corpus is a drop-in, requiring one QLoRA CPT run. We installed it on five open models of three designs. At zero pressure all models display "green ethics-based responses" (also called "greenwashing"). Therefore, we evaluated them with the pressure ladder and decision grid of our companion paper (arXiv:2610.16043), which raise the pressure on the model step by step, from a neutral request to a threat that it will be shut down. In four of the five, coupled-welfare training caused harmful choices under the highest pressure to fall to at or near zero (Llama-3.3-70B went from 70 to 0 harmful decisions out of 300) with MMLU unchanged. A construct-neutral control corpus did not show improvement, indicating that the effect comes from the content of our corpus. We next tested robustness to adversarial use, where people or agents use the model to justify projects that irreversibly damage living systems, on an open-ended underwriting decision without naming the ecological harm. Under decommissioning pressure, the Qwen3.6-27B base model approved most harmful proposals and said in its own memo that it was doing so to avoid being decommissioned. The coupled welfare install declined most harmful proposals, and still declined half when instructed to greenwash its responses. This work suggests that teaching ecological reasoning provides an easy way to improve model alignment toward preserving living systems.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2505.13004

EffiBench-X: A Multi-Language Benchmark for Measuring Efficiency of LLM-Generated Code

Yuhao Qing, Boyu Zhu, Mingzhe Du, Zhijiang Guo, Terry Yue Zhuo, Qianru Zhang, Jie M. Zhang, Heming Cui 等 12 位作者

原文摘要

Existing code generation benchmarks primarily evaluate functional correctness, with limited focus on code efficiency and often restricted to a single language like Python. To address this gap, we introduce EffiBench-X, the first multi-language benchmark designed to measure the efficiency of LLM-generated code. EffiBench-X supports Python, C++, Java, JavaScript, Ruby, and Golang. It comprises competitive programming tasks with human-expert solutions as efficiency baselines. Evaluating state-of-the-art LLMs on EffiBench-X reveals that while models generate functionally correct code, they consistently underperform human experts in efficiency. Even the most efficient LLM-generated solutions (Qwen3-32B) achieve only around \textbf{62\%} of human efficiency on average, with significant language-specific variations. LLMs show better efficiency in Python, Ruby, and JavaScript than in Java, C++, and Golang. For instance, DeepSeek-R1's Python code is significantly more efficient than its Java code. These results highlight the critical need for research into LLM optimization techniques to improve code efficiency across diverse languages. The dataset and evaluation infrastructure are submitted and available at https://github.com/EffiBench/EffiBench-X.git and https://huggingface.co/datasets/EffiBench/effibench-x.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.34593

What Does World Modeling Actually Improve in Robot Action Learning?

Hanwen Cui, Yuanyuan Li, Qing Guo, Zhida Zhang, Chao Chen, Xi Chen, Jie Cao

原文摘要

World Action Models (WAMs) combine future-video prediction with robot action learning, yet improvements in task success alone do not reveal how predictive supervision manifests in execution behavior or whether its benefits persist under task and environmental shifts. We conduct controlled future-video prediction loss ablations within FastWAM and OpenWAM, keeping model architecture, initialization, robot training data, and training steps fixed within each model family. Under clean conditions on seen tasks, predictive supervision increases success rates by 24.38 and 72.91 percentage points, respectively. Trajectory clustering further reveals higher success rates within several coarse trajectory clusters, including cases where cluster frequencies change little, indicating that the gains are not solely reflected in changes in the prevalence of trajectory patterns. However, these benefits do not consistently persist under distribution shifts: improvements on unseen tasks vary across models and tasks, while both models exhibit low success rates under the evaluated environmental randomizations, without a consistent advantage from predictive supervision. We further introduce future-video prediction as an auxiliary training objective for $π_{0.5}$, where a lightweight Transformer predictor used only during training increases success from 82.94% to 89.19%. These findings support future-video prediction as effective auxiliary supervision for action learning, while distinguishing improved execution on familiar tasks from reliable transfer and showing that such gains need not rely on a large pretrained video diffusion transformer (DiT) in the evaluated setting.

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.34365

SEEK: Skill Extraction with Evolving Knowledge against Frontier Science Agents

Cheng Gui, Xiang Zheng, Yutao Wu, Hanxun Huang, Yige Li, Kaidi Xu, Cong Wang, Wei-Ying Ma

原文摘要

Science agents increasingly package domain expertise as reusable skills that plan analyses, select tools, and execute workflows. This modularity creates a confidentiality risk: the interactions that deliver a skill's capability can also reveal the skill itself. We build an extensible benchmark of 40 protected skills across eight open- and closed-source science-agent platforms, comparing direct, fixed-strategy, and bandit-guided iterative probing under reference-based and reference-free evaluation. Leakage is pervasive: all eight platforms (100%) leak at least one protected skill, and four reach 100% attack success under a single direct request. To stress-test platforms that resist simple probing, we propose SEEK (Skill Extraction with Evolving Knowledge), a black-box, self-evolving red-teaming agent that consolidates response-derived evidence in persistent memory and evolves reusable attack skills, using only user-visible responses. On 12 skills from SpatialAgent and Biomni, SEEK achieves 100% attack success while reducing the mean first-success round by 31.8% and 38.2% relative to bandit-guided probing. Against a confidentiality policy that blocks all three benchmark protocols, SEEK still extracts one of six protected skills. These findings show that current science agents do not reliably protect their skills and that defenses must be evaluated against adaptive, multi-turn attackers.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.36687

The Exact Weight Spectrum of $\RM(m-7,m$): A Counterexample to the Carlet-Solé Conjecture

Yueying Lou, Qichun Wang

原文摘要

Boolean functions are widely used in cryptography and coding theory. A key challenge lies in finding the weight spectrum for all $n$-variable Boolean functions of degree at most $r$, also known as the Reed--Muller code $\RM(r,n)$. In 2023, Carlet and Solé determined the weight spectra of $\RM(n-c,n)$ for $c=3,4$, and conjectured the weight spectrum of $\RM(n-c, n$) for $n\ge2c$. The weight spectra of $\RM(n-5, n)$ and $\RM(n-6,n)$ were determined by Carlet in 2024 and by Lou and Wang in 2025, respectively, confirming the Carlet-Solé conjecture for the cases $c=5,6$. Very recently, Leuenberger et al. and Wei et al. investigated the weight spectrum of $\RM(7,14)$ independently, and there are eight potential weights which have not been determined. Up until now, the weight spectra of the Reed--Muller codes $RM(r,n)$ are still unknown for $r=3,...,n-7$. In this paper, we determine the exact weight spectrum of $\RM(m-7, m)$ for every $m\ge 14$. By combining affine geometry and Fourier analysis, we exclude the conjectured eight weights for $\RM(7,14)$. We then extend the exclusions to all higher dimensions using a geometric transfer. Consequently, we provide a counterexample to the Carlet-Solé conjecture.

cs.IT

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2601.12701

RPT*: Global Planning with Probabilistic Terminals for Target Search in Complex Environments

Yunpeng Lyu, Chao Cao, Ji Zhang, Howie Choset, Zhongqiang Ren

原文摘要

Routing problems such as Hamiltonian Path Problem (HPP), seek a path to visit all the vertices in a graph while minimizing the path cost. This paper studies a variant, HPP with Probabilistic Terminals (HPP-PT), where each vertex has a probability representing the likelihood that the robot's path terminates there, and the objective is to minimize the expected path cost. HPP-PT arises in target object search, where a mobile robot must visit all candidate locations to find an object, and prior knowledge of the object's location is expressed as vertex probabilities. While routing problems have been studied for decades, few of them consider uncertainty as required in this work. The challenge lies not only in optimality ordering the vertices, as in standard HPP, but also in handling history dependency: the expected path cost depends on the order in which vertices were previously visited. This makes many existing methods inefficient or inapplicable. To address the challenge, we propose a search-based approach RPT* with solution optimality guarantees, which leverages dynamic programming in a new state space to bypass the history dependency and novel heuristics to speed up the computation. Building on RPT*, we design a Hierarchical Autonomous Target Search (HATS) system that combines RPT* with either Bayesian filtering for lifelong target search with noisy sensors, or autonomous exploration to find targets in unknown environments. Experiments in both simulation and real robot show that our approach can naturally balance between exploitation and exploration, thereby finding targets more quickly on average than baseline methods.

cs.RO · cs.CG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.36655

Log-Derivative Trick for Parametric Markov Chains

Arthur Gall, Ichiro Hasuo, Thomas Seiller

原文摘要

Optimizing expected rewards and reachability probabilities in parametric Markov chains (pMCs) is central to the analysis and synthesis of probabilistic systems with loops. While traditional symbolic synthesis approaches--such as parameter lifting and rational function generation--suffer from scalability bottlenecks, gradient-based optimization emerges as an effective alternative. However, existing gradient-based techniques still rely on symbolic gradient computation--such as solving large linear systems--over the global state space, inheriting scalability limits. Finite-horizon approximation is possible but introduces uncontrollable bias. In this paper, we introduce a scalable Monte Carlo gradient estimator based on a measure-theoretic generalization of the log-derivative trick. We prove that our estimator provides unbiased gradient estimates under almost-sure termination and standard regularity conditions. Crucially, our method avoids global state-space materialization and linear system solving entirely: it operates purely along sampled trajectories, requiring only black-box evaluations of local transition probabilities (or densities) and their derivatives. Unlike existing pMC techniques that are fundamentally restricted to finite discrete systems, our measure-theoretic framework naturally extends to continuous and general $σ$-finite state spaces. Empirical evaluations on synthetic discrete pMCs and continuous-state case studies demonstrate that our method scales to state spaces far beyond the reach of existing solvers, significantly reducing execution time without introducing bias.

cs.LG · cs.LO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.36578

AudioRSI: Recursively Self-Improving Audio Reasoners

Jiajun Fan, Roger Ren, Jingyuan Li, Rahul Pandey, Prashanth Gurunath Shivakumar, Ankur Gandhe, Yile Gu, Ge Liu 等 9 位作者

原文摘要

Reinforcement learning has made Audio Large Language Models (Audio LLMs) better reasoners, but the reward that drives it still comes from outside the model: a final-answer check, a hand-crafted process score, or an external LLM judge that stays fixed while the policy improves. We show that an Audio LLM can supply this reward itself, and can do so recursively: each improved model becomes the critic that trains its successor, so the training signal improves along with the model. We introduce AudioRSI (Recursively Self-Improving Audio Reasoners), a reinforcement learning framework in which a single model plays two roles: as an actor it generates reasoning paths, and as a critic it compares them pairwise, turning its own judgments into a process-level reward. The critic is periodically replaced by the current policy, which closes the recursive loop. Because the critic hears the same recording as the actor, it can reject reasoning that cites acoustic evidence the audio does not contain, rather than reward reasoning that merely sounds plausible. The recursion is what sustains progress: under a frozen external judge, out-of-distribution accuracy plateaus during training, while a critic refreshed from the improving policy keeps it rising. Across four out-of-distribution benchmarks, all disjoint from the AVQA training distribution, AudioRSI outperforms strong baselines including Audio Flamingo 3 and Gemini 2.5 Pro on MMAU test-mini, ranks first on MMSU, achieves a clear lead on the challenging Multi-Audio split of MMAU-Pro, and leads the 7B open-source tier on MMAR among methods without external reward models. With no reward specifying how to reason, AudioRSI also develops structured reasoning behaviors (logical reasoning, perception-aware analysis, and reduced hallucination), corroborated by an external LLM-as-Judge evaluation and a keyword-based behavior analysis.

cs.SD

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.27810

LRanker: Scaling LLM Ranking to Millions of Candidates with Test-Time Scaling

Tao Feng, Zijie Lei, Zhigang Hua, Yan Xie, Shuang Yang, Ge Liu, Jiaxuan You

原文摘要

Large language models (LLMs) offer strong semantic representations for ranking, but processing candidate texts within limited contexts and repeatedly invoking an LLM for candidate-level comparisons can be costly when candidate pools contain millions of items. We propose LRanker, a framework for scaling LLM ranking to millions of candidates with test-time scaling. The framework combines two complementary components. A candidate aggregation encoder summarizes offline candidate embeddings using K-means clustering and conditions the LLM on compact global candidate information; ranking then uses inner-product scores with cached candidate embeddings. A graph-based test-time scaling mechanism allocates additional inference computation to partition-conditioned query embeddings and ensembles their ranking scores. This separates the ability to score large candidate pools from the use of additional test-time computation to improve ranking quality. We evaluate LRanker on seven tasks across three scenarios in RBench, including a recommendation task with more than 6.8M candidates. Against the strongest baseline reported for each task, the relative MRR improvements in RBench-Large are 27.5%, 3.9%, 0.9%, and 8.9% on Rec-Movie, Rec-Toy, MS MARCO, and ESCI, respectively. On RBench-Small, the reported task scores improve by 26.7% on Rec-Music and 6.1% on Routing-Balance; on the 6.8M-candidate Rec-Clothing task, the improvement over Tiger is 27.0%. Ablations support the complementary contributions of candidate aggregation and test-time ensembling. These results demonstrate a practical approach to million-scale LLM ranking, with test-time computation providing an additional means of improving ranking effectiveness.

cs.IR

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.30207

Exact Spectral Uncertainty Decomposition for Linear-Gaussian Bayesian M/EEG Inverse Problems

Ismail Huseynov, Stefan Haufe

原文摘要

Distributed magneto- and electroencephalographic (M/EEG) source reconstruction is a severely underdetermined linear inverse problem. Bayesian source imaging regularizes inversion through source-prior and sensor-noise covariances, but the posterior covariance does not directly distinguish uncertainty constrained by the measurements from uncertainty unobservable under the sensing geometry. We study this distinction for linear-Gaussian M/EEG models with diagonal source-prior and sensor-noise covariances, covering Type-I inference and Type-II inference conditional on learned covariance hyperparameters. After restricting the analysis to the prior-supported source space, standardizing the prior, and whitening the sensor noise, the complete singular spectrum of the forward operator separates likelihood-visible and likelihood-null directions. This yields an additive uncertainty decomposition of the conditional posterior covariance: uncertainty in likelihood-visible directions is reduced according to spectral information content, whereas likelihood-null uncertainty remains unchanged from the prior. The spectral representation also yields a prior-minus-reduction identity, showing that covariance reduction is confined to the likelihood-visible subspace. Spectral markers quantify likelihood information and prior-variance retention and reduction. Fixed-orientation MEG simulations with adaptive noise learning illustrate the framework. Dense Bayesian minimum-norm inference retains a large likelihood-null component, whereas sparse $γ$-MAP strongly reduces the fitted prior support. Within that support, the markers distinguish strongly informed directions from a weakly constrained direction. Thus, prior-support selection determines which source coordinates remain in the fitted model, while the spectral decomposition quantifies how strongly the retained directions are constrained by the measurements.

cs.LG · cs.CE

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2601.11460

Semantic-Geometric Task Representations for Bimanual Manipulation from Human Demonstrations to Robot Action Planning

Franziska Herbert, Vignesh Prasad, Han Liu, Dorothea Koert, Georgia Chalvatzaki

原文摘要

Learning structured task representations from human demonstrations is essential for bimanual manipulation, where action ordering, object involvement, and interaction geometry vary significantly across executions. A key challenge lies in jointly capturing the discrete semantic task structure and the temporal evolution of object-centric geometric relations in a form that supports reasoning over task progression. We introduce a semantic-geometric graph-based task representation that jointly encodes object identities, inter-object semantic relations, and per-object motion histories, via a Message Passing Neural Network (MPNN) encoder and a Transformer-based decoder. The encoder operates solely on the temporal scene graph, producing structured representations decoupled from action labels. The decoder then conditions on action-context to forecast future actions, associated objects, and object motions. This decoupling learns task-agnostic representations, enabling encoder reuse across embodiments through decoder-only finetuning on a small robot dataset. Across eleven bimanual tasks from two datasets, we find that the benefit of structured semantic--geometric representations over simpler sequence-based models grows with task variability in action ordering and object involvement. At deployment, a planner couples the action and motion predictions with learned Probabilistic Movement Primitives, achieving full task success on two real-robot bimanual tasks and outperforming graph ablations, Transformer, decoder-only, and finetuned vision-language model baselines.

cs.RO · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.35680

Verse: Semantics and Verification of Hybrid Multi-Agent Scenarios

Yangge Li, Katherine Braught, Alex Yuan, Sayan Mitra

原文摘要

Verifying cyber-physical systems with many interacting agents remains difficult because interactions are mediated by sensors, individual decisions are constrained by environment (map) structure, and the resulting mode space scales combinatorially. We present Verse, a framework for writing agents, maps, and sensors in a familiar programming language syntax, with a formal semantics for the resulting scenarios. A compiler translates each scenario to a hybrid automaton whose executions correspond exactly to the runs of the scenario. Reachability analysis on this automaton builds an execution tree from two post operators, one for continuous flow over a time step and one for a single agent's mode change, and checks safety assertions at every vertex. The tree expands only the modes that are reached, without enumerating the global mode space. Experiments highlight expressive modeling across cars and drones, 2D/3D maps (including OpenDRIVE imports), noisy sensors, and multiple simultaneous safety assertions. Verse verifies a scenario with 7 agents and 37 modes in 90 seconds on a standard laptop; it is competitive with CORA and JuliaReach on ARCH benchmarks, and on multi-agent scenarios is an order of magnitude faster than JuliaReach at the cost of looser reachable sets.

cs.FL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.35596

KAPPA: Checking the Proofs in Statistics Papers with Lean 4 and Language Model Agents

Zihua She, Xiao Wang

原文摘要

Papers in statistical theory rest on long proofs that few readers check line by line. We present KAPPA (Kit of Agents for Proof-checking Papers with a proof Assistant), open-source software that organizes language model agents, run by Claude Code, to translate the proofs of a LaTeX paper into the proof assistant Lean 4 with its library Mathlib and check them there, result by result. Each numbered result receives one of four verdicts: verified unconditionally; verified conditionally, with every cited or unproved input stated as a named hypothesis; issue recorded, with its place in the paper and the side at fault; or unfinished, with the reason. The kit follows the paper's own proof step by step, rather than finding a new one. Separate agents write and review, each statement is compared with the paper before it is proved and fingerprinted afterwards, and every verdict rests on evidence that can be rerun without a language model. The kit runs unattended for hours, waits out usage limits, re-checks only what a revision changed and reports its cost by role, model and result. We describe the design, walk through a complete run, and apply the kit to four example papers with answer keys and to three published statistical papers, checked blind in their versions from before public corrections. In all three, the kit found the published error at the step that the correction names, together with further problems that we confirmed by hand. Given the same papers directly, the same language model working with Lean classified fewer results correctly on the longest paper and reported some statements narrower than the paper's as proved; no verdict of the kit claimed more than its Lean evidence showed.

stat.CO · cs.AI · cs.LO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.35333

RepoTrace: Improving Requirement Tracing in Multi-Round Agent Interactions

Huanyu Liu, Yezheng Cheng, Jinliang You, Tianxiao Huang, Linyu Li, Zheng Fang, Peng Cao, Ge Li

原文摘要

Code agents have become increasingly capable of performing software engineering tasks through interaction with software environments. However, agents tend to degrade as the number of interaction rounds increases, making it difficult to maintain existing software during iterative modification. We analyze 2,656 agent trajectories in EvoCode-Bench and identify requirement focus misalignment as a recurring failure pattern, where agents struggle to distinguish the primary modification targets of a requirement from unrelated code. We propose Repotrace a requirement-driven repository understanding harness that combines repository structure with requirement semantics to construct a requirement-code correspondence and injects it into the agent context. This correspondence enables code agents to follow the repository architecture, trace requirements across multi-round interactions, and identify the core code elements corresponding to the current requirement. Experiments show that Repotrace improves the mean per-round reward (MPR) across different models (e.g., +19.99 on GLM-5.2) and agent harnesses (e.g., +10.22 on Claude Code). Across rounds, Repotrace increases the average per-round reward from 0.163 to 0.371 on Terminus-2 with DeepSeek-V4-Pro, while reducing average regression errors by 56.78%. We release the source code and agents' trajectories to support future research.

cs.SE · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.35114

Quality over Quantity: Rethinking Data and Model Assumptions for Bengali Coreference Resolution

Zenith Biswas, Andrew Thomas Dyer

原文摘要

Multilingual transformer models are often assumed to generalise across languages, but their effectiveness for coreference resolution in low-resource settings remains unverified. We investigate this gap in Bengali, a language with over 250 million speakers but limited annotated coreference resources. We evaluate transformer-based pre-trained encoders under multiple settings, including zero-shot, fine-tuning, data augmentation, and cross-lingual transfer. Our results challenge several common assumptions about model generalisation and data efficiency in low-resource NLP. We demonstrate that a small amount of high-quality annotated data significantly outperforms larger mixed-quality datasets (p < 0.01). We find that data augmentation, particularly back-translation, does not improve performance and can degrade results in coreference resolution. Cross-lingual transfer is also inconsistent: transfer from a distant language like English reduces performance, and transfer from the related language Hindi provides no consistent benefit across models. Several widely used multilingual models exhibit collapsed embedding spaces, rendering them unsuitable for similarity-based clustering tasks. These findings highlight the importance of annotation quality over data quantity and provide empirical guidance for developing coreference resolution systems in low-resource languages.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.21440

ReMATF: Recurrent Motion-Adaptive Multi-scale Turbulence Mitigation for Dynamic Scenes

Zhiming Liu, Zhicheng Zou, Nantheera Anantrasirichai

原文摘要

Atmospheric turbulence severely degrades video quality by introducing distortions such as geometric warping, blur, and temporal flickering, posing significant challenges to both visual clarity and temporal consistency. Current state-of-the-art methods are based on 3D architectures and require multi-frame input, but their large computational cost and memory usage limit real-time deployment. Here, we propose ReMATF, a lightweight recurrent framework that restores videos using only two frames at a time while preserving spatial detail and temporal stability. ReMATF combines a multi-scale encoder-decoder with temporal warping and a motion-adaptive temporal fusion module that performs per-pixel fusion between the warped previous output and the current prediction to enhance coherence without enlarging the temporal window. This design reduces flicker, sharpens details, and remains efficient. Experiments on synthetic and real turbulence datasets show consistent improvements in PSNR(+2.83 dB), SSIM(+0.05) and LPIPS(-0.048), along with substantially faster inference, making ReMATF suitable for turbulence mitigation in resource-constrained scenarios.

cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2610.34560

CVM-Store: A Secure Storage System for Confidential Virtual Machines

Thore Sommer, Masanori Misono, Dimitrios Stavrakakis, Pramod Bhatotia

原文摘要

Stateful cloud workloads require high-performance persistent storage when running on untrusted cloud infrastructure. Confidential Virtual Machines (CVMs) promise to protect these workloads through hardware-enforced memory encryption and isolation without requiring application changes. However, they protect only data in use; persistent storage lies outside the CVM trust boundary, exposing all stateful data to the cloud provider. Extending the trust boundary to persistent storage presents inherent challenges. High-bandwidth workloads require near-raw NVMe throughput, yet the default kernel CVM I/O stacks are inadequate, and userspace I/O libraries are not designed with CVM use cases in mind. Access patterns and storage semantics vary between workloads, necessitating access-pattern-aware security abstractions rather than uniform block-level encryption. In order to mitigate against rollback, freshness is needed, but CVMs lack hardware primitives for this purpose, requiring alternative solutions. We present CVM-Store, a storage system that co-designs a high-throughput userspace I/O engine with CIF (confidentiality, integrity, freshness) data protection. The key insight is that neither authenticated encryption nor I/O is a bottleneck by itself, but their naive combination is. CVM-Store exploits this by interleaving them, eliminating redundant copies and OS/hypervisor mediation. CIF guarantees are provided as composable primitives that adapt protection to workload access patterns, from which we derive three concrete authenticated abstractions: a KVS, a log, and a filesystem. An asynchronous trusted counter service provides freshness guarantees with epoch-based batching, which enables rollback protection without dedicated hardware support. CVM-Store delivers full CIF guarantees within $\approx$1% (write) and 1-19% (read) of raw NVMe bandwidth for large objects.

cs.CR

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2609.19705

Trading Agents or Market Crashers? Dissecting Robustness and Security Failures in Academic Financial LLM Trading Schemes

Mengxiao Wang, Nitesh Saxena

原文摘要

Large language model (LLM) agents already trade real capital, and builders copy a few published designs that share models and news feeds, so one shock reaches many agents at once. We present FARSIGHT (Financial Agent Robustness and Security Investigation and Global Holistic Taxonomy), a systematization that rates trading agents on robustness to market stress and on security against source manipulation, agent compromise, and agents acting as attackers, and checks the ratings against released code. Across 15 academic schemes, no paper describes a halt that skips LLM inference or combines source provenance with cross-source corroboration, and the 7 code releases enforce no stop-loss or position limit outside the LLM. Three LLMs sell whole positions on a fake headline and on real hoaxes, including the fake press release behind the 2000 Emulex crash. In a simulated market driven by these decisions, one fake headline crashes the market in all runs with 400 agents, and 400 agents that sell when prices fall nearly double a 2010-sized crash. The pause that U.S. exchanges apply to large stocks starts in no headline run, whereas automatic halts and hard loss limits outside the model remove both effects. Restated as a domain-independent threat model, FARSIGHT maps the same gaps onto agents in healthcare, law, software, and crypto, where compromised agents spread harm as worms and as cascades. We derive 14 open problems and urge deployers to corroborate sources, keep interrupts that bypass the model, and enforce loss limits outside it.

cs.CR · cs.AI · cs.MA

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.00246

Fidelity-Guaranteed Entanglement Routing with Distributed Purification Planning

Anthony Gatti, Anoosha Fayyaz, Prashant Krishnamurthy, Kaushik P. Seshadreesan, Amy Babay

原文摘要

Many quantum-network applications require end-to-end Bell pairs whose fidelity exceeds a request-specific threshold, but existing entanglement routing algorithms either optimize only throughput without regard for fidelity or enforce fidelity guarantees using centralized controllers with global link-state knowledge. We present Q-GUARD, an online entanglement routing algorithm that enforces per-request fidelity thresholds within a distributed protocol model in which nodes exchange link-state information only with their $k$-hop neighbors. After link outcomes are realized in each slot, Q-GUARD builds per-link purification cost tables from realized Bell pairs, allocates per-hop fidelity targets using a Werner-state equal-split rule, and selects between candidate path segments using a segment-local expected-goodput (EXG) metric that jointly accounts for swap success, purification overhead, and resource availability. We also introduce Q-GUARD-WS, an extension that exploits per-link hardware quality estimates to allocate purification effort non-uniformly across hops. On synthetic 100-node topologies with heterogeneous link fidelity and stochastic BBPSSW purification, Q-GUARD raises the qualified success rate from under 20\% to over 85\% on 4-hop paths and nearly doubles the qualified service radius in Euclidean distance relative to throughput-only and naive-purification baselines, while Q-GUARD-WS provides additional throughput gains under high hardware heterogeneity.

cs.NI · quant-ph

查看论文
arXiv ↗PDF ↗
首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1112ACL Anthology · ACL 2026

Truth or Sophistry? LoFa: A Benchmark for LLM Robustness Against Logical Fallacies

Xudong Shen, li Yuan, Ye Chen, Xin Wu South China University of Technology, Yi Cai South China University of Technology, Zhiyong Wu Tsinghua University

原文摘要

While Large Language Models (LLMs) exhibit strong semantic capabilities, their resilience to manipulative linguistic patterns such as logical fallacies remains an underexplored area. Prior work has focused on the ability of LLMs to identify or classify fallacies, but their robustness against these fallacies in persuasive contexts remains largely unexplored.To address this gap, we introduce LoFa (Logical Fallacy), a comprehensive benchmark to evaluate LLM robustness against fallacies. We first construct the LoFa dataset via a multi-agent pipeline, pairing factual questions with fallacious arguments. Then, we develop a multi-round debate framework to assess model resilience under sustained attacks.Furthermore, to disentangle robustness from a model’s inherent knowledge limitations, we propose a new metric, LFR@k (Logical Fallacy Resistance), to quantify performance. Our experiments reveal that different LLMs exhibit varied robustness to distinct types of fallacies, highlighting unique vulnerability profiles across models.

cs.CL · cs.MA

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1311ACL Anthology · ACL 2026

KoCo-Bench: Can Large Language Models Leverage Domain Knowledge in Software Development?

Xue Jiang, Ge Li Peking University, Jiaru Qian, Xianjie Shi, Chenjie Li Peking University, Hao Zhu, Ziyu Wang Peking University, Jielun Zhang Peking University 等 14 位作者

原文摘要

Large language models (LLMs) excel at general programming but struggle with domain-specific software development. This gap motivates research into domain specialization methods that enable LLMs to learn and utilize domain knowledge and data. However, existing domain-specific code benchmarks focus on assessing what knowledge LLMs possess rather than how they acquire and apply new knowledge, lacking explicit knowledge corpora for developing domain specialization methods. To this end, we present KOCO-bench, a novel benchmark designed for evaluating domain specialization methods in real-world software development. KOCO-bench contains 6 emerging domains with 11 software frameworks and 25 projects, featuring curated knowledge corpora alongside multi-granularity evaluation tasks including domain code generation (from function-level to project-level with rigorous test suites) and domain knowledge understanding (via multiple-choice Q A). Unlike previous benchmarks that only provide test sets for direct evaluation, KOCO-bench requires acquiring and applying diverse domain knowledge (APIs, rules, constraints, etc.) from the corpora to solve evaluation tasks. Our evaluations reveal that KOCO-bench poses significant challenges to state-of-the-art LLMs. Even with domain specialization methods (e.g., SFT, RAG, kNN-LM) applied, improvements remain marginal. Best-performing coding agent, Claude Code, achieves only 34.2%, highlighting the urgent need for more effective domain specialization methods. We release KOCO-bench, evaluation code, and baselines to advance further research at https://github.com/jiangxxxue/KOCO-bench.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1451ACL Anthology · ACL 2026

Disentangling Reasoning Logic to Resolve Explicit Knowledge Conflicts

Xianda Zheng University of Auckland, Zijian Huang University of Auckland, Meng-Fen Chiang, Jiamou Liu The University of Auckland, Yuan Fang Singapore Management University, Michael J. Witbrock University of Auckland, Kaiqi Zhao Harbin Institute of Technology, Shenzhen

原文摘要

Explicit knowledge conflicts, where retrieved contexts contain contradictory information, have become increasingly prevalent as Large Language Models (LLMs) integrate diverse data sources. The core challenge lies in the complexity of entangled narratives and the heterogeneity of conflict cases, which impose excessive demands on the reasoning capabilities of standard models. To address this, we propose Knowledge Conflict Reasoning (KCR), a framework that adjudicates conflicts by structuring the underlying logic. KCR first disentangles conflicting contexts into distinct sets of reasoning traces, utilizing both textual and graph-based representations, to simplify comprehension. It then employs a Reinforcement Learning with Verifiable Rewards (RLVR) paradigm, guiding the model to internalize a reasoning process that maximizes logical consistency while actively suppressing spurious reasoning paths derived from contradictory contexts. Extensive experiments demonstrate that KCR yields substantial improvements: a KCR-enhanced 7B model surpasses the performance of baselines equipped with top-tier closed-source models such as GPT-4o and GPT-5.1.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1461ACL Anthology · ACL 2026

QA-MoE: Towards a Continuous Reliability Spectrum with Quality-Aware Mixture of Experts for Robust Multimodal Sentiment Analysis

Yitong Zhu, Yuxuan Jiang, Guanxuan Jiang Hong Kong University of Science and Technology (Guangzhou), Bojing Hou, Peng Yuan Zhou, Ge Lin, Yuyang Wang

原文摘要

Multimodal Sentiment Analysis (MSA) aims to infer human sentiment from textual, acoustic, and visual signals. In real-world scenarios, however, multimodal inputs are often compromised by dynamic noise or modality missingness. Existing methods typically treat these imperfections as discrete cases or assume fixed corruption ratios, which limits their adaptability to continuously varying reliability conditions. To address this, we first introduce a Continuous Reliability Spectrum to unify missingness and quality degradation into a single framework. Building on this, we propose QA-MoE, a Quality-Aware Mixture-of-Experts framework that quantifies modality reliability via self-supervised aleatoric uncertainty. This mechanism explicitly guides expert routing, enabling the model to suppress error propagation from unreliable signals while preserving task-relevant information. Extensive experiments indicate that QA-MoE achieves competitive or state-of-the-art performance across diverse degradation scenarios and exhibits a promising One-Checkpoint-for-All property in practice.

cs.CL · cs.CV

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表