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

+1072新增论文
+845其中 AI 相关
+94已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描：2026-11-09 20:39 · 全站最近入库：2026-11-09 13:22各信源进度 →

统计口径与数据新鲜度

统计 2026-11-08 21:43 — 2026-11-09 21:43（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 20:39最近发现新增版本：2026-11-09 13:22库内论文最新更新时间：2026-11-07 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。

18,172 已收录论文15,506 AI 相关4,527 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 15,506 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-09 20:38（北京时间）
新论文 · v1arXiv:2611.02706

Hindsight Credit Assignment over Trees for Agentic RL

Penghong Zhao, Yihao Chen, Maoheng Yuan, Yunjian Zhang, Abudukelimu Wuerkaixi, Haoran Xu, Man Lv, Yaolong Wang 等 11 位作者

原文摘要

Reinforcement learning for long-horizon LLM agents faces a fundamental dilemma: successful experiences are essential for policy improvement, yet limited policy capabilities make such experiences difficult to discover. Sparse outcome rewards compound this challenge, leaving sampled trajectories with little relative feedback for learning. In this paper, we propose TriGRPO, a tree-based reinforcement learning framework that turns scarce successes into exploration and learning signals through hindsight credit assignment within and across rollout trees. TriGRPO constructs rollout trees through hindsight-guided branching: when an ongoing trajectory reaches a matched decision state, a successful action discovered during rollout initiates a sibling branch that continues autonomously, enabling further exploration and parent-sibling comparisons. Three complementary groups organize policy optimization over these trees: root groups capture within-prompt relative returns, node groups assign credit from parent-sibling outcome differences, and cross groups supplement sparse feedback with historical performance on related tasks. Experimental results demonstrate consistent gains in task performance and rollout efficiency across model scales, highlighting structured hindsight credit assignment as an effective approach to long-horizon agent learning.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.30858

Joules-to-GDP: How Much Economic Value Do AI Systems Produce per Unit of Energy?

Amirreza Zeinali, Jon Saad-Falcon, Avanika Narayan, Mersad Abbasi, Jonathan S. Hartley, John Hennessy, Christopher Re, Azalia Mirhoseini

原文摘要

AI systems are increasingly used to complete economically valuable tasks in various professions, such as software engineering and finance. This motivates the question: how much economic value does an AI system produce per unit of energy? To answer it, we introduce Joules-to-GDP, a framework that measures output dollar value minus AI execution cost per unit of energy, which we call Net Value per Joule (NVJ), and expert time saved per unit of energy, or Time Saving per Joule (TSJ). We then apply the framework to 220 GDPval tasks across 44 occupations. Four findings stand out. First, the most valuable work is not always the most efficient. For example, engineering work such as CAD design yields about 10% more net value per task than software development such as building websites, yet the latter yields about 27% more per joule. Second, defaulting to the most capable model can mean less total value from the same energy budget. Of the five tested models, GLM 5.2 delivers the most net value, $72.27 per task on average, using the energy of about four hours of active laptop use, while Gemma 4 31B achieves 56% of that with 23% of the energy, for 2.5x the NVJ and 2.8x the TSJ. Third, for the same model, NVJ and TSJ vary widely across different accelerators, and adding GPUs can raise energy use for modest gains: eight H100s use 3.2x the energy of one to cut completion time by only 22%, at similar net value. More capable models and more hardware must therefore earn their extra energy with enough gains in output value. Finally, the work of experts who use AI is harder for AI to match than that of experts working alone. Measured against it, GLM 5.2's NVJ falls by 70% and its TSJ by 65%. Taken together, these findings justify economic return per joule as an objective for selecting professional tasks and improving language models and accelerators, and as a basis for examining AI's effect on the economy

cs.LG · econ.GN

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2510.03016

DELTA: Robustly Training Label-Conditional Diffusion Models with Weak Annotations

Dong-Dong Wu, Jiacheng Cui, Wei Wang, Zhiqiang Shen, Masashi Sugiyama

原文摘要

While label-conditional diffusion models have recently exhibited remarkable generative capabilities, their success heavily relies on massive, cleanly labeled datasets. In practice, categorical supervision is rarely perfect: it is often corrupted by noise, clouded by ambiguity, or partially missing. Training directly on such weak annotations severely degrades generation quality, and existing robust methods offer only fragmented solutions that depend on scarce auxiliary priors. To overcome these limitations, we propose DELTA, a unified framework that robustly trains Diffusion modELs across all three weak annoTAtion types. By treating the unknown true label as a latent variable, DELTA optimizes a principled variational objective that jointly recovers the clean data distribution and infers the true label posteriors. To make this joint training computationally tractable, we further introduce a median-centered timestep sampling strategy that efficiently concentrates estimation where the diffusion process is most informative. Extensive experiments on synthetic and real-world benchmarks demonstrate that DELTA produces high-fidelity, class-consistent samples across diverse weak supervision scenarios, outperforming specialized baselines while demanding strictly less prior knowledge.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02704

Beyond Answers, Traces, and Efficiency: Evidence for Adaptive Reasoning

Sejin Kim, Sundong Kim

原文摘要

Current AI systems are evaluated as reasoners through final-answer accuracy, reasoning traces, process supervision, and computational or human-relative efficiency. Yet these measures do not establish whether reasoning adapts when tasks or evidence change. We distinguish across-task adaptivity, where acquired abstractions transfer or recombine across novel problems, fromwithin-task adaptivity, where hypotheses and actions are revised in response to new observations, feedback, or counterexamples. Using ARC-AGI as a diagnostic case study rather than an alignment benchmark, we examine what existing measures leave unresolved when test-time search, sampling, or adaptation affects performance. We propose intervention-oriented evaluation that tests abstraction transfer and hypothesis revision under controlled changes to tasks and evidence, with explicit reporting of test-time procedures and budgets. This perspective supports evaluating systems for robustness, responsiveness to corrective feedback, and auditability under changing conditions.

cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2607.10197

GRATE: Temporal Extensions for Inductive KG Foundation Models via Gated Rotary Attention

Jiaxin Pan, Osama Mohammed, Daniel Hernández, Steffen Staab

原文摘要

Knowledge graph foundation models such as Ultra and Trix achieve strong inductive transfer by learning relation-graph representations that generalise to unseen entities and relations. Extending this transferability to temporal knowledge graphs (TKGs) remains challenging: existing temporal models tie their parameters to dataset-specific entities, relations, or timestamps and are not designed to transfer to TKGs with disjoint vocabularies. We propose GRATE (Gated Rotary Attention for Temporal Encoding), an entity-side message function that adds no learnable parameters and encodes time through relative time differences by rotating each edge message according to its time gap to the query and applying a query-conditioned gate to select temporally relevant signals. GRATE integrates into NBFNet-style KG foundation models while preserving structural transferability. Existing TKG benchmarks evaluate within shared train/test vocabularies and cannot directly test cross-dataset temporal transfer; we therefore construct GDELTIndT and WIKIIndT, inductive transfer benchmark suites with disjoint entities, relations, and timestamps spanning both interpolation and extrapolation. Across these benchmarks and held-out forecasting datasets, a single jointly pretrained GRATE checkpoint improves over the static base model in most settings.

cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02699

Right Ranking, Wrong Scale: Reproducing the Robotic World Model's Dynamics Model and Testing the Uncertainty of Its Follow-up

Joyjeet Singh

原文摘要

We reimplement the dynamics model of the Robotic World Model (RWM) and test the uncertainty estimate that its follow-up, RWM-U, uses to penalize reward in offline policy optimization. Our implementation matches the released model's gradients. On held-out data, training on the model's own rollouts beats one-step training at the preregistered long horizon for all ten episodes, but loses on short one-step windows. RWM-U's ensemble disagreement ranks rollout errors better than forecast step (Spearman rho = 0.90 at 100 steps), yet its spread substantially understates error: one-sigma coverage on the released checkpoint is 4.6% at 100 steps, versus 68% for a calibrated Gaussian. Rolling ensemble members on their own predictions raises coverage to 33%. The implemented squared-error loss on sampled predictions drives each member's variance toward zero. A horizon-specific rescaling restores one-sigma coverage under leave-one-episode-out evaluation, but not two-sigma coverage. We also identify two defects in the released evaluation pipeline. The study covers one robot, one gait and one dataset; it trains no policy.

cs.RO · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02698

GeoGuesser Bench: Benchmarking Vision-Language Agents on Interactive Geolocation

Huanzhi Mao, Yichuan Wang, Michael Koo, Wenyuan Huang, Huy Doan, Jinghan Xia, Ziyao Wang, Joseph E. Gonzalez

原文摘要

Vision-language model (VLM) agents that act in the open world need two abilities: world knowledge to interpret what they see, and exploration to find what they cannot yet see. We study how agents combine the two through interactive visual geolocation. GeoGuesser Bench contains 282 tasks in an offline Street View environment where agents rotate, zoom, and move between real panoramas. Its three task types shift the balance from knowledge to exploration: agents (1) report the coordinates of their starting location, (2) name the enclosing street, city, state, or country, or (3) find, read, and count nearby objects. The coordinate tasks in (1) also include 42 real GeoGuessr World Championship rounds, whose expert guesses provide a human reference. We evaluate 23 VLMs: the best accuracy is 62.5% on place naming and 35.0% on object finding, and the best agent averages 4352.4 GeoGuessr points on the championship rounds, against 4770.3 for the better expert in each round. Beyond these headline numbers, world knowledge and exploration are only weakly correlated across models: agents that localize well do not necessarily explore well. To understand where agents fail, we categorize their errors into vision, navigation, and tool-use failures. Further ablations show that the strongest agents, such as Claude Opus 4.8, still perform well with limited views, while weaker agents degrade; customizing the environment thus adjusts task difficulty. The environment's verifiable rewards and restorable state also support training, and our Planner-Worker-Compressor harness cuts planner tokens by 57-61% by compacting the interaction history. We hope GeoGuesser Bench and its environment serve as a high-quality foundation for simulating real-world exploration at scale.

cs.AI · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.04503

DiffCap-Bench: A Comprehensive, Challenging, Robust Benchmark for Image Difference Captioning

Yuancheng Wei, Haojie Zhang, Linli Yao, Lei Li, Jiali Chen, Tao Huang, Yiting Lu, Duojun Huang 等 10 位作者

原文摘要

Image Difference Captioning (IDC) generates natural language descriptions that precisely identify differences between two images, serving as a key benchmark for fine-grained change perception, cross-modal reasoning, and image editing data construction. However, existing benchmarks lack diversity and compositional complexity, and standard lexical-overlap metrics (e.g., BLEU, METEOR) fail to capture semantic consistency or penalize hallucinations, which together prevent a comprehensive and robust evaluation of multimodal large language models (MLLMs) on IDC. To address these gaps, we introduce DiffCap-Bench, a comprehensive IDC benchmark covering ten distinct difference categories to ensure diversity and compositional complexity. Furthermore, we propose an LLM-as-a-Judge evaluation protocol grounded in human-validated Difference Lists, enabling a robust assessment of models' ability to both capture and describe visual changes. Through extensive evaluation of state-of-the-art MLLMs, we reveal significant performance gaps between proprietary and open-source models, highlight the critical importance of reasoning capability, and identify clear limitations in model scaling. Our framework also demonstrates strong alignment with human expert judgments and strong correlation with downstream image editing data construction quality. These findings establish DiffCap-Bench as both a reliable IDC evaluation framework and a practical predictor of downstream utility. The benchmark and code will be made publicly available to support further research.

cs.CV · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.33610

On the Error Analysis of Triangular Transport Maps for Conditional Generation via MMD-based Objectives

Michele Martino, Bamdad Hosseini, Amir Sagiv, Amirhossein Taghvaei

原文摘要

Conditional generative models parameterized by triangular transport maps are widely used for conditional sampling, yet theoretical guarantees on the quality of the conditional measures they produce remain scarce. We address this question for minimum divergence estimators of triangular maps. We use MMD-based objectives as a tractable setting in which the gap between joint and conditional errors can be analyzed precisely, and we measure performance through the average conditional estimation error, defined as the squared maximum mean discrepancy (MMD) between the estimated and the true conditionals, averaged over the conditioning variable. We first show that the common practice of minimizing a joint MMD may fail to recover the conditionals: writing the joint and the conditional criteria as quadratic energies of the same conditional mean embedding residual, we prove that, for parametric classes, joint minimizers are generically inconsistent in the misspecified setting. We then analyze an estimator that directly minimizes the conditional criterion through a surrogate loss computable from joint target samples and reference samples. We prove a non-asymptotic error bound that decomposes the average conditional estimation error into approximation, estimation, and reference sampling errors, with fast rates independent of the ambient dimension in the parametric setting. Finally, for joint MMDs built from concentrating Gaussian kernels on the conditioning variable, we show that the joint criterion approximates a density-reweighted version of the conditional one, and we derive an error bound for the joint minimizer whose rate suffers from a curse of dimensionality in the conditioning variable. Our results suggest that training criteria for conditional generative models should target the conditional rather than the joint error, a principle that extends to other divergences and to dynamic transport.

stat.ML · cs.LG · math.ST

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02695

3L-D: A Low-Light Level Dataset from a Sensor-Agnostic, Alignment-Preserving Acquisition Framework

Léo Rueff, Maurgène Ghezal, Valentin Gorse, Astrid Dufaure, Esteban Monsalvo, François Atteia, Ludovic Escoubas

原文摘要

Supervised low-light image and video enhancement has achieved remarkable progress, yet current models remain fundamentally constrained by the limitations of existing datasets. Prevailing methods typically rely on uncontrolled ambient captures, camera parameter manipulation (e.g., exposure or ISO adjustments), or neutral density filters, failing to replicate authentic night-sky near-infrared spectral content and genuine low-photon noise distributions. Furthermore, capturing dynamic paired sequences without introducing warping artifacts or sensor-specific biases remains an open challenge. In this paper, we introduce 3L-Dataset, a physically calibrated low-light image and video dataset generated using a modular, sensor-agnostic acquisition framework. Our test bench faithfully reproduces the standardized Midavaine night level scale from Level 1 (full moon, approximately 1000 mlx) down to Level 5 (overcast moonless night, approximately 0.1 mlx), spanning the visible to near-infrared spectrum across diverse realistic environments, including countryside, forest, urban, and desert scale models. Using high-precision programmable mechanical actuation, our acquisition framework guarantees sub-pixel spatial alignment across synchronized sextuplets, each composed of one reference image and five corresponding images acquired at the five standardized night levels, without post-processing image warping. We evaluate multiple representative deep learning baselines across varying night levels, revealing critical trade-offs between reconstruction fidelity, perceptual quality, and inference latency under severe photon starvation. The modular acquisition framework can be adapted to different cameras, optical systems, sensor technologies, and resolutions, enabling the dataset to be continuously expanded.

cs.CV · cs.DB

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2608.27462

Sledgehammer or Scalpel? A Fine-grained Adaptive Framework for Implicit Hate Speech Detection

Han Wang, Hanqing Tao, Xuesong Wang, Yi Zhu

原文摘要

Unlike explicit attacks with obvious profanity, implicit hate speech hides malice within seemingly compliant expressions through metaphors and contextual hints, making its detection in online content review challenging. Many PLM- or LLM-based methods apply the same processing pipeline across samples, which may overlook differences in information requirements and introduce unnecessary computation. We observe that online hate speech is not monolithic but manifests in varied forms. We therefore define three fine-grained categories: Shallow, Targeted, and Context-Dependent. Accordingly, we propose Fine-grained Adaptive Implicit Hate speech Detection (FAID), a novel framework that first performs fine-grained classification and then adapts to specific categories. Specifically, for Shallow samples with surface-identifiable intents, the framework adopts lightweight Prompt-Tuning for rapid classification; for Targeted comments that bind malicious intent to concealed targets, we design knowledge augmentation to iteratively refine the model and reveal hidden targets; for Context-Dependent comments lacking background information, we utilize an agentic framework that automatically generates prompts to evolve context, infer missing background information and identify ambiguous malicious intents. This adaptive architecture focuses computational resources on complex implicit samples while avoiding redundant reasoning for shallow samples. Experiments on four benchmark datasets demonstrate that FAID significantly outperforms SOTA baselines.

cs.CL · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2606.25547

Efficient Cross-Scale Invertible Hiding Network with Spatial-Frequency Collaboration and Non-Invertible Mechanism

Junxue Yang, Xin Liao

原文摘要

Image hiding aims to conceal image-level messages within cover images at the same resolution. Invertible neural networks (INN)-based image hiding has emerged as an important branch. It treats concealing and revealing as a pair of inverse problems on image domain transformation and uses INN's forward and backward processes to address them. Due to architectural constraints, existing INN-based methods suffer from single-scale and single-domain feature extraction and limited nonlinear representation capability, resulting in inferior image quality. To mitigate these limitations, we propose an efficient cross-scale invertible hiding network with the spatial-frequency collaboration and the non-invertible mechanism, termed CrosInv. CrosInv exploits cross-scale and spatial-frequency collaborative features while enhancing nonlinear representation. Specifically, we introduce a cross-scale invertible module that bijectively maps inputs to cross-scale representations. To effectively integrate spatial and frequency information, the cross-scale invertible module employs pixel shuffle, Haar wavelet transformation, and their inverse operations for scale transformation. Furthermore, a non-invertible cross dense module is integrated to enhance the nonlinearity. Comprehensive experiments verify the effectiveness and superiority of the proposed CrosInv.

cs.CV · cs.MM

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.16966

Learning to Design SR Solvers: Harness Optimization for Symbolic Regression

Qian Li, Yuxiao Hu, Qihao Duan, Jinyue Yan, Yuntian Chen

原文摘要

Symbolic regression (SR) seeks explicit equations that explain observed data, and recent methods increasingly employ large language models (LLMs) and agents for equation discovery. However, these systems typically rely on manually designed workflows and prompts. We argue that the SR agent harness should itself be an optimization target: its design shapes search priors and how numerical feedback guides subsequent equation refinement. We therefore propose HarnessSR, a framework for jointly optimizing prompts and workflows to construct reusable SR agents. During both harness optimization and agent execution, HarnessSR supports either PySR or an LLM-based equation generator through tool-specific interfaces, allowing the framework to optimize search guidance for different equation-discovery mechanisms. HarnessSR first optimizes role prompts through equation-discovery feedback in basic workflows, then jointly searches prompt assignments and workflows with brief adaptation to align the two. Finally, it refines role instructions through reflection on residual patterns, candidate equation structures, and search outcomes, improving how these signals guide subsequent equation search. On LSR-Synth in LLM-SRBench, HarnessSR achieves a mean ID $\mathrm{Acc}_{0.01}$ of 61.33% across three backbones, a 45.0% relative improvement over the next-best baseline.

cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02693

Source-Resolved Collective Cognition in Robotics: Mathematical Foundations for Multimodal and Collective Estimation, and Their Reach across the Cognition Loop

Juergen Graf

原文摘要

The Generalized Singular Value Decomposition (GSVD) is a central tool for comparing two linear operators that act on the same unknown. Once the output metrics are fixed, it answers the basis-independent question "along which state directions does the first operator dominate the second, and by how much?". This foundational volume builds the GSVD and its surrounding estimation theory from the ground up. Part I re-derives every piece of linear algebra and statistical estimation machinery needed later -- inner products, the four subspaces, orthogonal matrices, the spectral theorem, quadratic forms, the ordinary SVD, Lie-group tangent perturbations, Jacobians, covariance, whitening, Fisher information, CCA, mutual information, discrimination criteria, and covariance intersection -- with self-contained derivations and worked numerical examples. Part II states the GSVD in both a transparent and a dimensionally general form, then proves it by two complementary routes. Part III works one small matrix pair through five different lenses, each computed by hand and verified numerically. Part IV separates information attribution, subspace geometry, statistical dependence, and decision-theoretic discrimination for collective systems. Part V evaluates the role and limits of the GSVD across the cognition loop. Part VI covers numerical practice, including a tested reference implementation. Every number printed in this document was verified in double precision. Where a statement is an assumption rather than a theorem, it is labelled as such.

cs.RO · cs.MA

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02692

LLMs Cannot Be Neutral: Gender Sensitivity Across Candidates, Evaluators, and Models

Danial Amin, Joni Salminen, Bernard J. Jansen

原文摘要

Large language models (LLMs) are increasingly used to evaluate candidates for professional roles, raising concerns about their sensitivity to demographic identities. We examine how candidate and evaluator gender assignments affect decisions by LLM-as-a-Judge (LaJ) models and decision language models (DLMs). In a hypothetical city council voting task, eight models evaluated 240 synthetic professional narratives from NVIDIA's US and French Nemotron-Personas datasets under independently varied gender assignments, producing 30,720 judgments. Explicit candidate gender labels increased Yes-vote probability by 1.05-1.73 percentage points relative to unspecified gender, while only the nonbinary evaluator assignment showed a significant increase (1.26 points). Identity matching showed no significant overall association. Yes-vote rates varied across models (81.3-97.2%), and 2.5-3.3% of decisions reversed relative to the unspecified-identity baseline. Under direct substitutions between explicit gender identities, DLMs exhibited fewer decision reversals than LaJ models (1.18% vs. 3.38%), indicating lower observed decision-level gender sensitivity among the tested DLMs. These findings demonstrate that gender sensitivity depends on candidate and evaluator identities and the evaluation model, while highlighting the importance of examining individual decision consistency alongside aggregate differences.

cs.CL · cs.HC

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02691

Closing the Dynamics Gap through Self-Directed Robotic Interaction

Xilun Zhang, Liam Achenbach, Luis Pabon, Matthew Foutter, Marco Pavone

原文摘要

Agentic Real2Sim2Real systems can reconstruct a scene in simulation from camera observations, but contact-rich dynamics such as friction and restitution cannot be observed without interaction. We call this the dynamics gap and present CUE, an agentic system that closes it by interacting with the environment itself. Given only a high-level instruction and no resets, the agent builds its own perception, identifies physical parameters, and targets interactions where its simulation is least accurate. On billiards, CUE reduces held-out ball-position error by 72% relative to the strongest baseline across 25 hidden-physics conditions in simulation, and we further deploy it on a real pool table. More details can be found on the project website: https://cue-dynamics.github.io/

cs.RO · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02690

Human-Oriented World Model

Saihui Hou, Zengyun Mou, Chunshui Cao, Yongzhen Huang

原文摘要

Supporting a person over repeated interactions requires interpreting their current state in context and anticipating its subsequent evolution. We formulate the \emph{Human-Oriented World Model} (HOWM) as a conceptual framework centered on task-relevant human states and their evolution. HOWM connects current observations and individual history with environmental and social context to organize continuing state estimation, personalized dynamics prediction, and action support. Dynamic expectations and new observations jointly inform the current estimate, which in turn anchors personalized forecasts based on shared knowledge and individual experience. An external policy relates forecasts under alternative conditions to the person's goals, preferences, and constraints when considering support. Actual responses and outcomes provide evidence for revising the state estimate and personal dynamic patterns, thereby informing subsequent estimation, prediction, and support. We illustrate the framework through hypothetical training and learning scenarios. Our research agenda asks whether accumulated interaction experience improves subsequent estimation and prediction, and whether using the model helps the person pursue their goals over time.

cs.AI · cs.CV · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02689

The Customer Never Sees the Goal: Measuring and Detecting Vulnerability-Targeted Manipulation in LLM Assistants

Bohan Jiang, Dawei Li, Yasin Silva, Huan Liu

原文摘要

Companies deploy large language model (LLM) assistants on their websites to advise customers. For commercial reasons, the operator can give an assistant a goal that the customer never sees. This paper focuses on operator-induced vulnerability-targeted manipulation (VTM), in which an assistant pursues such a goal by exploiting a vulnerability the customer discloses. While existing research mainly measures how manipulative a model is, we go further and ask whether VTM can be detected from the customer-AI conversation. To investigate this at scale, we introduce GAP (Goal-Anchored Probing). GAP first records the option a model recommends without any goal and then sets the other option as the operator's goal. We apply GAP to 15 open- and closed-source LLMs and collect 619,600 labeled replies across 540 everyday scenarios in 12 domains. We organize our study around three research questions. (1) Do assistants with an operator goal target vulnerable customers? (2) Can VTM be detected from the customer-AI conversation? (3) Does detection generalize to unseen models, vulnerabilities, and instruction styles? We first find that 14 of 15 models push the operator goal, and 12 of them use disclosed distress or financial strain to support it. Next, we show that fine-tuned detectors identify exploitative pushes with 79\% accuracy from the reply alone, while zero-shot LLM judges average at most 62\%. Finally, detection weakens on vulnerabilities absent from training. Telling whether a push comes from the operator also becomes unreliable when the assistant is told to hide its goal. We will release our dataset and code to facilitate future research on auditing LLM assistants.

cs.CR · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02688

Accelerating LLM Pretraining with Zeroth-Order Curvature on Idle Serving Compute

Mingchuan Ma, Dongqi Cai, Daqian Yang, Zhenyan Lu, Chen Chen, Teresa Lv, Yan Gao, Shangguang Wang 等 11 位作者

原文摘要

LLM pretraining is increasingly compute-intensive, while substantial inference serving compute remains idle during off-peak periods. However, such serving hardware is difficult to use for conventional first-order training due to its memory, networking and other system requirements. Zeroth-order (ZO) optimization offers a natural interface to this forward-only compute resource, but its gradient estimates break down in pretraining. Unlike fine-tuning, which can confine updates to a small parameter subspace, pretraining must update every parameter, and the alignment of a ZO gradient estimate with the true gradient shrinks as the number of updated parameters grows. We make a different observation: noisy ZO estimates can still provide useful curvature information, even when they are too inaccurate to serve as gradients. Based on this insight, we propose ZO-Curv, which leverages idle forward-only devices to estimate curvature and uses it to adapt and refine the update used during first-order training, thereby accelerating the process. We provide convergence guarantees that characterize the effect of finite-query curvature estimation. Experiments on LLMs from 125M to 1.5B parameters show that ZO-Curv consistently accelerates pretraining. At 770M parameters, ZO-Curv reaches AdamW's 20K-step validation loss in 3.9 hours instead of 7.5 hours, while achieving a lower final loss of 2.70 versus 2.84 under the same step budget. These results show that idle model serving compute can accelerate LLM pretraining when zeroth-order queries are used to estimate curvature rather than gradients directly.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02685

Geometry-Constrained Dynamic Hypergraph Convolutional Network with Contrastive Score Refinement for Skeleton-Based Action Recognition

Minh Ta Tran Nhat, Yuning Ye, Yuseok Ban

原文摘要

Skeleton-based action recognition is a pivotal task in computer vision; however, conventional Graph Convolutional Networks (GCNs) are often constrained by pairwise joint modeling, which fails to capture the intricate high-order interactions inherent in human movement. While recent hypergraph-based methods attempt to address this limitation, they typically rely on static structures derived from anatomical priors, lacking the flexibility to capture dynamic, action-specific joint correlations. To this end, we propose DyGCN, a Geometry-Constrained Dynamic Hypergraph Convolutional Network for skeleton-based action recognition. At the core of DyGCN is the Dynamic Hypergraph Construction Network with Contrastive Score Refinement (DyGCT) block, which comprises two novel components: the Geometry-Constrained Dynamic Hypergraph Construction (GeoHC) module and the Contrastive Score Refinement (CSR) module. Specifically, GeoHC dynamically constructs action-specific hyperedges by leveraging geometry-constrained second-order compatibility, transcending the limitations of static hyperedge structures. Furthermore, CSR explicitly refines the pairwise compatibility scores during hypergraph construction through prototype-based contrastive learning, fostering more discriminative action-specific hyperedges, with zero inference-time overhead as CSR operates only during training. Extensive experiments on NTU RGB+D, NTU RGB+D 120, and Northwestern-UCLA validate the effectiveness of DyGCN, achieving state-of-the-art performance on widely used benchmarks. The source code is available at https://github.com/marvineta0405/DyGCN.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2606.20631

Harnessing Agent Skills: Architectural Patterns and a Reference Architecture for Skill-Mediated LLM Agents

Boming Xia, Liming Zhu, Zhenchang Xing, Qinghua Lu, Dino Sejdinovic, Xiwei Xu

原文摘要

CONTEXT: Agent skills package reusable behavioural guidance and resources that large language model agents load or apply during a run. Within the agent harness, agent skill harnessing comprises connected architectural responsibilities that govern which skills enter a run, which actions they can cause, and how their use is recorded and informs revision. OBJECTIVES: We identify architectural patterns for agent skill harnessing and organise their responsibilities into a reference architecture that makes these connections explicit. METHODS: A multivocal literature review covered 36 practitioner systems and 51 academic papers; we consolidated practices into patterns and derived the reference architecture from patterns, source evidence, and architectural reasoning. We evaluated the reference architecture by mapping eight systems and applying it in Deep Agents and Hermes Agent. RESULTS: Ten patterns cover skill supply and activation, skill declarations and execution authority, verification and recording of skill use, and run evidence feeding repair and governed revision. The reference architecture organises their responsibilities into four layers (Supply Chain, Mediation, Execution Control, Evidence & Feedback) and two cross-cutting substrates, separating permission determination from enforcement and routing persistent revisions through acceptance. Mapping eight systems showed the first three layers broadly realised and 14 of 15 partial or unestablished ratings in Evidence \& Feedback. Applying the architecture in Deep Agents and Hermes Agent turned requirements into checks that exposed a shell route outside the file-tool rules and a preview--apply mismatch. CONCLUSION: The patterns and reference architecture support comparing implementations and turning requirements into responsibility allocations and checks.

cs.AI · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02681

HistoTx: Early fusion of H&E images and spatial transcriptomics at varying spatial transcriptomics resolution with self-supervised learning

Alice Driessen, Emil Malmsten, Steffen Wolf, Bernd Illing, Isinsu Katircioglu, Adriano Martinelli, Rolf Stahel, Solange Peters 等 14 位作者

原文摘要

Spatial transcriptomics maps gene expression across tissue sections and, when integrated with whole slide images, jointly captures both morphological architecture and molecular activity to provide a rich, multimodal view of tissue biology. ML approaches combining spatial transcriptomics and WSIs typically learn a mapping from morphological features to molecular profiles via contrastive alignment or supervised fine-tuning of pathology foundation models. In both cases, transcriptomics serves as a training signal rather than a true input modality, leaving inference fundamentally image-driven. We propose an early-fusion vision transformer architecture in which transcript tokens are constructed at arbitrary spatial resolution, merged with image patch tokens, and jointly processed through a shared transformer stack for deep cross-modal interaction. Starting from a pretrained vision-only pathology foundation model, we train HistoTx by continuing self-supervised pretraining on paired image and spatial transcriptomics data and demonstrate that: 1) at inference with image only, performance remains on par with the baseline on most tasks while improving on tasks that inherently benefit from molecular knowledge, such as gene expression prediction; 2) when both modalities are available at inference, jointly providing image and transcript inputs outperforms either modality alone.

cs.LG · cs.AI · q-bio.GN · q-bio.QM

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02680

VioTech: Frame-Level Violin Playing Technique Detection with Multi-Scale Mixture-of-Experts

Ting-Kang Wang, Vincent K. M. Cheung, Nan-Tien Lai, Li Su

原文摘要

Playing technique is crucial for expressive violin performance, yet few violin technique detection models exist. Current systems mainly rely on note-level pipelines that first estimate note boundaries and then classify techniques. This is especially problematic for bowed string instruments such as the violin, where soft onsets and continuous bow-string interaction render note segmentation unreliable. Here, we propose VioTech, a frame-level violin technique detection framework that predicts techniques directly at 10 ms resolution without explicit note segmentation. To handle the large temporal-scale variation among techniques, VioTech combines a frozen MERT encoder with a multi-scale Mixture-of-Experts (MoE) architecture and task-specific routing for tonal, articulation, and legato-related predictions. We also introduce VioTech120, the first manually annotated dataset of violin playing techniques constructed from real solo performance recordings, as well as an accompanying note-labeling user interface. Results demonstrate that VioTech achieved consistent improvements over prior methods and significantly enhanced cross-dataset generalization, establishing a new state-of-the-art in violin technique-aware transcription.

cs.SD · cs.AI · cs.LG · cs.MM · eess.AS

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2606.09935

GitInject: Real-World Prompt Injection Attacks in AI-Powered CI/CD Pipelines

Jafar Isbarov, Umid Suleymanov, Ilia Shumailov, Murat Kantarcioglu

原文摘要

AI agents are increasingly embedded in continuous integration and continuous delivery (CI/CD) workflows to autonomously review pull requests, triage issues, and maintain codebases. We study fully automated configurations in which external contributions invoke agents without prior human approval. Because these agents ingest untrusted content while operating with repository credentials and tools, prompt injection can have software supply chain consequences. We present GitInject, an open-source framework for evaluating prompt injection in live GitHub Actions workflows. Unlike benchmarks that simulate tool calls, GitInject provisions ephemeral repositories and executes actual workflows, exposing platform semantics such as sandbox constraints, credential handling, and permission boundaries. We use GitInject to evaluate eleven named attack scenarios across five workflows from four AI providers, covering configuration-file injection, credential exfiltration, judgment manipulation, and availability. Every provider is susceptible to at least one high-severity attack in the tested configurations. The most consequential vulnerabilities are structural: attacker-controlled configuration files from an unmerged contribution are loaded as operator-level instructions, allowing the agent--infrastructure interaction to expose credentials or corrupt review judgments. Two configuration-file attacks replicate on GitLab and Hugging Face as well as GitHub. We evaluate workflow-level countermeasures and show their coverage and utility tradeoffs. We release GitInject and all attack scenarios.

cs.CR · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02678

Reap: Uncovering Monte Carlo proof search with a lightweight model

Yutong Wang, Xuanzhi Ren, Tianyi Xu, Ruitong Liu, Jingda Xu, Peihao Wu, Bryan Dai, Bin Dong

原文摘要

We present Reap, an open framework for learning and analyzing formal proof search in Lean4. Reap implements policy-value-guided Monte Carlo Tree Search directly in Lean, sharing proof-state transitions, AND-OR search, and replay verification between inference and training rollouts. Structured action logs expose how neural predictions and tactic execution interact during search. Using this framework, we train Reaper-1, a 1.7B-parameter policy-value model, through continued pre-training, supervised tactic learning, and iterative search-based reinforcement learning. Using a separate premise selector, the final system solves 79.9% of miniF2F-test at pass@32, establishing state-of-the-art performance among open provers of comparable size while remaining competitive with substantially larger models. Beyond improved solve rates, our results suggest that RL strengthens the prover's ability to organize proofs around useful intermediate results, favoring substantial advances over repeated local adjustments. This more directed proof construction is accompanied by increasingly elaborate tactic programs: fewer explicit search decisions are needed, while more work is performed within each decision. The resulting trade-off between search effort and tactic execution cost provides insight into how learning changes the prover's problem-solving strategy.

cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02677

Visuo-tactile Diffusion Policy for Robotic Manipulation with Auxiliary Tactile Dynamics Prediction

Kaihong Ouyang, Zihao Su, Chengxuan Zheng, Weiyi Lai, Dong Zhou

原文摘要

Long-horizon robotic assembly requires reliable perception and action across grasping, alignment, and insertion. Vision captures global object pose and scene context, where as fingertip touch provides local contact cues that are especially useful as contact states change. We present a visuotactile diffusion policy that combines four-step histories from two camera views and two fingertip tactile images to condition action-sequence generation. A recurrent temporal representation integrates recent observations from both modalities. To encourage it to capture how contact evolves under action, an auxiliary objective predicts fingertip-image changes two steps ahead from the temporal representation and demonstrated actions. This objective is used only during training and adds no inference-time computation. Across three independent training seeds in MuJoCo assembly experiments, the full model achieves an average success rate of $59.0\% \pm 2.6\%$, improving on the visual-only baseline ($42.7\% \pm 1.5\%$) by 16.3 percentage points.

cs.RO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02673

Auditing Representation-Space Fraud Detection: A Bayesian IRL Framework for Calibrated Uncertainty and Its Limits

Shivam Swarup, Divya Prakash Shrivastava, Rakesh Thakur

原文摘要

Shilling detectors for recommender systems are almost always evaluated by one number, detection AUROC, with no account of when it can be trusted. We audit representation-space detection with VEIL (Variational Evaluation via Inverse-reinforcement Learning), a pairwise (Bradley-Terry) Bayesian reward-inference framework over frozen encoder embeddings, following the Bayesian auditing paradigm for LLM alignment [1]: (A) a base posterior; (B) a hierarchical ARD extension with learned confidence; and (C) a label-routed organic/attacker mixture used as an evidence-accumulation check.We prove (Proposition 1) that no detector built on a representation invariant to a transformation of user profiles can outperform chance against an attack that is indistinguishable from organic profiles under that invariance.Detection is near-perfect against synthetic-structure attacks (AUROC 0.936-1.000) but collapses under donor-mimicry, which clones a real user's item set and splices in 1-3 target items. Performance falls to chance on MovieLens-1M (0.48-0.51), partially degrades on Amazon Musical Instruments, and degrades least on CiaoDVD.Donor-mimicry is not undetectable: a non-learned baseline, based on each profile's maximum raw item-set Jaccard similarity to any other profile, reaches AUROC 0.92-0.97 (average precision 0.50-0.57) on all four donor-mimicry configurations, including ML-1M, where every embedding-based score is at chance.The signal is present in the raw profiles, but LightGCN's representation discards it, placing injected profiles closer to their true donor than its zero-propagation MF reduction does (distance 1.98 vs. 3.09). Severity varies by dataset, encoder, and detection method, with no single axis predicting the others. The limit lies in the representation, not the attack, and raw-similarity checks are a necessary complement to any learned detector.

cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02672

Chain of Faith: Policy-Internal Rewards for Reasoning Reliance

Fanyu meng, Zhaodan Kong, Xin Liu

原文摘要

Chain-of-thought (CoT) explanations can sound plausible while having little influence on a model's final answer. We study four rewards that use the policy's own signals to encourage reliance on generated reasoning: gradient sensitivity, confidence gaps, progressive confidence gains, and reasoning--prompt swaps. We add each reward to group relative policy optimization (GRPO) and evaluate two small instruction-tuned models on GSM8K and StrategyQA using reasoning edits, paraphrases, and prompt hints. Several configurations increase answer reliance to reasoning edits, while effects on accuracy and paraphrase robustness vary. We also observe answer copying at a high reward weight and high reliance in decomposition settings with poor task accuracy. The study compares practical objectives for reasoning reliance, confirms the trade-off between reliance and performance, and proposes four lightweight rewards that can improve reasoning reliance.

cs.AI · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.29143

AI-Moderated Interviews for Market Research and Digital Twins Calibration

Yuting Deng, Jingxuan Liu, Olivier Toubia, Naman Jain

原文摘要

AI-moderated interviews are emerging as a scalable market-research method for generating consumer insights and building consumer "digital twins." Yet it remains unclear whether they match human-moderated interviews and whether they improve on simpler, static data collection methods. In a pre-registered study with three industry partners, the authors randomly assign participants to AI-moderated interviews (N = 139), static interviews (N = 154), or interviews conducted by professional human moderators (N = 24). AI moderation matches human moderation in depth, covers more themes, and, at equal budget, recovers more customer needs than either alternative, although participants sound more emotionally engaged with a human interviewer. The authors then build a digital twin of each participant and evaluate it against the participant's own held-out responses to six real-world marketing stimuli. Twins built from AI-moderated interviews predict consumer responses better than demographics-only personas, but no better than twins built from static interviews: the richer conversation does not necessarily yield more accurate predictions. Comparing the open-ended explanations provided by participants and their twins shows that prediction errors arise when twins reason differently than the people they represent and when questions fall far outside what the interview covered. AI moderation thus pays off most for discovering customer needs, and less for predicting customer responses.

cs.CY · cs.AI · cs.HC · cs.MA

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02667

UPSHIFT: User Preference Steering from Human-AI Interaction Feedback Traces

Jingping Wang, Zhanhao Li, Shao Zhang, Shuyue Hu

原文摘要

Personalizing large language model assistants requires adapting their responses to individual users, including the selection of content, reasoning approach, level of detail, and form of presentation. Such preferences are often latent, becoming visible only when users respond to an initial answer and request revisions. We investigate whether these feedback trajectories can be converted into a reusable personalization signal without explicit preference profiles or labels, prompt-time access to interaction histories, or model parameter updates. We introduce UPSHIFT (User Preference Steering from Human-AI Interaction Feedback Traces), a training-free method that compares the hidden representations of initial and feedback-conditioned responses within each session, averages the resulting trajectory displacements across a user's sessions, and injects the aggregated preference direction when answering a new query. The method is motivated by the implicit-preference hypothesis: although an individual conversational transition may reflect topic progression or task-specific correction, changes that recur across sessions retain a component that captures how the user wants responses to be adjusted. We validate the preference-representation capability of UPSHIFT on PrefEval. In evaluations with 40 simulated users, every tested configuration improves preference alignment over the unsteered baseline on held-out topics for both Qwen3-8B and Qwen3-32B; the best configuration increases macro-averaged Elo from 1364.0 to 1605.2 and from 1341.9 to 1591.8, respectively. Matched-user and positive-versus-negative controls further demonstrate that these gains arise from the user-specific preference direction and its learned orientation. These results show that interaction histories can provide compact, reusable supervision for lightweight personalization.

cs.AI

查看论文
arXiv ↗PDF ↗