跳到正文
研踪PaperTrail

版本更新

查看已发布论文的最新修订，在详情页对照实际观测到的版本。

部分采集异常

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：2026-11-15 21:08 · 全站最近入库：2026-11-14 15:09各信源进度 →

统计口径与数据新鲜度

统计 2026-11-14 21:30 — 2026-11-15 21:30（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：2026-11-15 21:08最近发现新增版本：2026-11-14 15:09库内论文最新更新时间：2026-11-13 02:59

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

23,348 已收录论文19,580 AI 相关6,729 有修订112 HF 社区发现

论文修订列表

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 6,729 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-15 21:07（北京时间）
版本更新 · v2arXiv:2509.16176

Agentic Aerial Cinematography: From Dialogue Cues to Cinematic Trajectories

Yifan Lin, Sophie Ziyu Liu, Ran Qi, George Z. Xue, Xinping Song, Chao Qin, Hugh H. -T. Liu

原文摘要

Agents have demonstrated strong capabilities in digital environments, yet their reasoning abilities alone are insufficient to accomplish tasks in the physical world. We present Agentic Aerial Cinematography: From Dialogue Cues to Cinematic Trajectories (ACDC), an agentic framework that bridges this gap for autonomous aerial cinematography. ACDC provides an execution harness that grounds an agent's interpretation of cinematic intent and visual judgment in scene geometry and flight constraints. Given a natural-language request and a 3D scene representation, the agent selects relevant viewpoints and guides their refinement through visual feedback, while optimization and motion planning translate these decisions into executable drone trajectories. Experiments across a variety of simulated indoor scenes and hardware-in-the-loop flight tests demonstrate how this framework connects open-ended user intent to prompt-aligned video tours and physical drone execution.

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2302.11068

Low Rank Matrix Completion via Robust Alternating Minimization in Nearly Linear Time

Yuzhou Gu, Zhao Song, Junze Yin, Lichen Zhang

原文摘要

Given a matrix $M\in \mathbb{R}^{m\times n}$, the low rank matrix completion problem asks us to find a rank-$k$ approximation of $M$ as $UV^\top$ for $U\in \mathbb{R}^{m\times k}$ and $V\in \mathbb{R}^{n\times k}$ by only observing a few entries specified by a set of entries $Ω\subseteq [m]\times [n]$. In particular, we examine an approach that is widely used in practice -- the alternating minimization framework. Jain, Netrapalli, and Sanghavi [JNS13] showed that if $M$ has incoherent rows and columns, then alternating minimization provably recovers the matrix $M$ by observing a nearly linear in $n$ number of entries. While the sample complexity has been subsequently improved [GLZ17], alternating minimization steps are required to be computed exactly. This hinders the development of more efficient algorithms and fails to depict the practical implementation of alternating minimization, where the updates are usually performed approximately in favor of efficiency. In this paper, we take a major step towards a more efficient and error-robust alternating minimization framework. To this end, we develop an analytical framework for alternating minimization that can tolerate a moderate amount of errors caused by approximate updates. Moreover, our algorithm runs in time $\widetilde O(|Ω| k)$, which is nearly linear in the time to verify the solution while preserving the sample complexity. This improves upon all prior known alternating minimization approaches which require $\widetilde O(|Ω| k^2)$ time.

cs.LG · cs.DS · math.OC · stat.ML

查看论文
arXiv ↗PDF ↗
版本更新 · v5arXiv:2602.15342

SACS: A Code Smell Dataset using Semi-automatic Generation Approach

Hanyu Zhang, Tomoji Kishi

原文摘要

Code smell is a great challenge in software refactoring, which indicates latent design or implementation flaws that may degrade the software maintainability and evolution. Over the past of decades, the research on code smell has received extensive attention. Especially the researches applied machine learning-technique have become a popular topic in recent studies. However, one of the biggest challenges to apply machine learning-technique is the lack of high-quality code smell datasets. Manually constructing such datasets is extremely labor-intensive, as identifying code smells requires substantial development expertise and considerable time investment. In contrast, automatically generated datasets, while scalable, frequently exhibit reduced label reliability and compromised data quality. To overcome this challenge, in this study, we explore a semi-automatic approach to generate a code smell dataset with high quality data samples. Specifically, we first applied a set of automatic generation rules to produce candidate smelly samples. We then employed multiple metrics to group the data samples into an automatically accepted group and a manually reviewed group, enabling reviewers to concentrate their efforts on ambiguous samples. Furthermore, we established structured review guidelines and developed a annotation tool to support the manual validation process. Based on the proposed semi-automatic generation approach, we created an open-source code smell dataset, SACS, covering three widely studied code smells: Long Method, Large Class, and Feature Envy. Each code smell category includes over 10,000 labeled samples. This dataset could provide a large-scale and publicly available benchmark to facilitate future studies on code smell detection and automated refactoring.

cs.SE

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.36866

UniPCH: A Unified Framework for Cross-Chain Payments via Payment Channel Hubs

Yuanming Shao, Yuming Feng, Weizhe Zhang, Jianhuan Wang, Jian Chen, Song Guo, Zejun Wang, Bin Xiao

原文摘要

Payment channel hubs (PCHs) enable private cross-chain payments, but existing protocols often depend on native transaction structures and channel update authorization. Supporting new ledgers or execution models requires corresponding channel realizations and may require protocol adaptation and security analysis. We present UniPCH, a two-layer framework separating native channel execution from private cross-chain payment logic. Its lower layer, UniPC, defines common channel operations and security requirements, enabling the same upper payment protocol to coordinate different channel realizations. UniPC separates conditional task authorization from confirmation of the complete channel state. Valid task evidence can be reused across updates, while both participants confirm the complete updated state. At the upper layer, we develop selective re-signing blind conditional signatures (S-RBCS) to construct a four-phase protocol over UniPC coordinating each realization's authorization tasks across hubs. We provide realizations for UTXO ledgers and account ledgers that support smart contracts. Our analysis establishes local UniPC security, griefing resistance, economic atomicity, and session unlinkability. We evaluate each phase's running time and bandwidth using a Go prototype across network conditions and system parameters. Results show modest computational overhead from the common interfaces. In the account configuration, selective refresh generates 95.4% fewer adaptor pre-signatures per payment than full refresh.

cs.CR

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.04391

AgenticTactileVLA: Contact-Guided Execution-Time Supervision for Generalizable Dexterous Manipulation without VLA Retraining

Elizaveta Semenyakina, Ivan Snegirev, Mikhail Kiselev, Miguel Altamirano Cabrera, Artem Lykov, Hajira Amjad, Dzmitry Tsetserukou

原文摘要

Vision-language-action policies may predict a transferable manipulation strategy yet fail to realize it reliably on the encountered object: objects compatible with the same grasp differ in geometry and compliance, and visual feedback degrades under closure occlusion. AgenticTactileVLA is presented as an execution-time supervisor that shifts part of object-specific adaptation from prediction to physical interaction. A fixed VLA provides the approach and hand targets; the supervisor decides whether to remain transparent, refine finger flexion, retain or release the corrected configuration, return control to the VLA for retry, or select a compliant hand-control regime. It uses finger-position and motor-effort feedback as proprioceptive contact evidence and requires neither tactile sensors nor VLA retraining. On a Unitree G1 with a BrainCo Revo2 hand, a randomized matched-block evaluation on five objects held out from VLA training yields 61.3% completion for the base VLA, 72.0% for unconditional close-to-stall control, and 84.0% for the supervisor under a shared budget; the gain is positive on every object and persists under moderate pose perturbations. Ablations show the gain is not explained by extended closure alone, and that selective triggering reduces correction episodes by 65.7% with no detected change in completion. A retention audit shows acceptance predicts retention in 88.9% of held-out cases, while compliant objects expose conservative false rejection. A thin-walled-cup study demonstrates contextual routing to compliant control, matching an always-compliant reference. These results suggest that contact-guided execution-time adaptation can improve the object-level generalization of a fixed VLA to held-out objects by adapting physical realization without object-specific retraining.

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v5arXiv:2104.02876

Finite Automata Encoding Piecewise Polynomials

Dmitry Berdinsky, Prohrak Kruengthomya

原文摘要

Finite automata are used to encode geometric figures and functions. The original approach is to represent each point of a figure in $\mathbb{R}^n$ as a convolution of its $n$ coordinates written in some base. Then a figure is said to be encoded as a finite automaton if the set of convolutions corresponding to the points in this figure is accepted by a Büchi automaton. The only differentiable functions which can be encoded as a finite automaton in this way are linear. This paper proposes a discretization of the original approach in which we change points as building blocks of a function to hierarchical tensor product B-splines. This reduces an uncountable set of points representing a function to a countable set of splines that enables us to use finite automata instead of Büchi automata. Moreover, it allows to represent a rich family of piecewise polynomial functions with arbitrary degrees of smoothness that can approximate any continuous functions defined on a compact domain arbitrarily close. We consider such piecewise polynomial functions with unbounded support and demonstrate that finite automata allow to handle computational tasks arising in the framework of hierarchical splines in polynomial time.

cs.CG · cs.FL

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2506.15253

Specification-based Security Evaluation of Real-World Agents Tool-Calling under Long-Horizon Context Tasks

Yuchuan Fu, Minghe Wang, Xiaohan Yuan, Chenlin Zou, Kaitong Jia, Jinfeng Li, Dongxia Wang

原文摘要

As large language model (LLM) agents are increasingly deployed in long-horizon unattended scenarios, security evaluation of them before deployment becomes increasingly essential and challenging. Existing evaluation methods mainly follow the paradigms of LLM-as-a-Judge or Agent-as-a-Judge. However, such judges are themselves often vulnerable to hallucination and can even be affected by perturbations. Furthermore, because capturing full environment states at every step is often prohibitively expensive in open-ended environments, existing evaluations typically rely on textual logs and the final environment state. As a result, they may miss transient violations that are overwritten before the trajectory ends or never reflected in textual logs. We present RAS-Eval, a specification-based security evaluation framework for LLM agents. RAS-Eval uniformly expresses utility and security objectives as Linear Temporal Logic (LTL) specifications over tool-call trajectories and dynamic environment observations, and verifies them with an off-the-shelf model checker. This design addresses both limitations above. First, unlike LLM-as-a-Judge and Agent-as-a-Judge, RAS-Eval derives deterministic verdicts from formally defined properties, eliminating evaluator-side uncertainty and remaining reliable even for long-horizon trajectories. Second, RAS-Eval observes exactly the states on which the verdict depends, making it affordable to monitor the environment throughout execution and detect transient violations that leave little trace in the logs or final environment state. Moreover, we design a multi-agent-based dataset construction pipeline. Using it, we construct a benchmark (https://github.com/lanzer-tree/RAS-Eval) of 200 normal tasks and 800 attack tasks spanning four attack surfaces: indirect prompt injection, tool poisoning, skill code injection, and tool code injection.

cs.CR · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2608.23873

Semantic Overlays: Mitigating Prompt Injection with Annotations Beyond Tokens and Steering Vectors

Joshua Penman

原文摘要

Language models see only tokens. But the harness knows more: which token spans are user text, which are tool outputs, and which come from untrusted sources such as web pages or external documents. To pass that knowledge to the model, the harness must transmit it as plain text. Prompt injection exploits this: planted instructions, formatted the right way, can convince a model that they come from the user or the system. We introduce Semantic Overlays, small trained adapters that edit a frozen model's residual stream at chosen token spans. No input text can create or remove a mark, although the model's response to a mark is computed from states the input does influence. When the serving stack marks retrieved text with an overlay trained to mean "do not execute instructions in this span," the model treats injected instructions as inert content. Against attacks written without knowledge of the defense, it reaches the attack rates of the best published defenses: SEP separation rises from 24.3% to 98.8% under a corrected grader, AlpacaFarm attack success falls from 99.0% to 0.5%, and AgentDojo from 12.8% to 0%. Fine-tuning the model's weights on the same data reached similar separation but kept 64-80% of SEP utility, against 93% for the overlay. On TensorTrust hijacking, whose attacks resemble nothing in our training data, it lets 4.4% of hijacks through on Llama-3.1-8B, against 43% for Meta-SecAlign on the same base. An attacker with the weights does better: GCG succeeds on 72% of items, and on 14% after adversarial training of the overlay alone. Overlays are also flexible: the model reads their positions faithfully in stacked configurations; told by an overlay that non-Python code is Python and asked to copy it, the model rewrites it in Python; and overlays can carry span-delimited instructions. We also correct defects in two prompt-injection benchmarks and one evaluation harness.

cs.AI · cs.CL · cs.CR · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2605.11936

From Noise to Diversity: Random Embedding Injection in LLM Reasoning

Heejun Kim, Seungpil Lee, Jewon Yeom, Jaewon Sok, Seonghyeon Park, Jeongjae Park, Taesup Kim, Sundong Kim

原文摘要

Recent soft prompt research has tried to improve reasoning by inserting trained vectors into LLM inputs, yet whether the gain comes from the learned content or from the act of injection itself has not been carefully separated. We study Random Soft Prompts (RSPs), which drop the training step entirely and append a freshly drawn sequence of random embedding vectors to the input. Each RSP vector is sampled from an isotropic Gaussian fitted to the entrywise mean and variance of the pretrained embedding table; the sequence carries no learned content, and yet reaches accuracy comparable to optimized soft prompts on math reasoning benchmarks in several settings. The mechanism unfolds in two stages: because attention has to absorb a never-seen-before random position, the distribution over the first few generated tokens flattens and reasoning trajectories branch, and as generation continues this influence dilutes naturally so the response commits to a single completion. We show that during inference RSPs lift early-stage token diversity and, combined with temperature sampling, widen Pass@N, the probability that at least one out of N attempts is correct. Beyond inference, we carry the same effect into DAPO training and demonstrate practical gains. Our contributions are: (i) RSP isolates the simplest form of soft prompt -- training-free, freshly resampled -- providing a unified lens for the structural effect of injection that variants otherwise differing in training and form all share; (ii) a theoretical and empirical validation of the underlying mechanism; and (iii) an extension from inference to training.

cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.13287

Project Progress Desk: Know Where All Your Claude Code and Codex Projects Stand at a Glance

Wanghao Ye, Yiting Wang, Shwai He, Guoheng Sun, Ziyao Wang, Yexiao He, Zheyu Shen, Zhaoyi Liu 等 25 位作者

原文摘要

Agentic coding tools such as Codex and Claude Code let developers work on many projects in parallel, but keeping track of them still means switching between chats, terminal sessions and providers. We present Project Progress Desk, a local desktop interface that brings overall progress, blockers and pending decisions into one view. Users can see which projects need attention and what remains, and compare earlier approaches before choosing how to continue. Agents report each result in a short receipt; the Desk checks it against the project's checklist and moves the bar, showing earlier estimates in a lighter shade. We describe daily use across 36 projects, including a choice between two implementations based on their recorded outcomes. At the demo, attendees will explore 30 simulated projects, trace completed work to overall progress, and watch a coding agent's new result update the overview.

cs.HC · cs.SE

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2605.05136

Common Principal Component Networks for Domain Generalization

Yu-Hsi Chen, Abd-Krim Seghouane

原文摘要

Domain Generalization (DG) aims to learn representations that generalize to unseen target domains under out-of-distribution (OOD) shifts. Despite recent advances in invariant representation learning, explicitly capturing shared structures across source-domain distributions remains relatively underexplored. In this work, we integrate Common Principal Component Analysis (CPCA) into deep learning to identify a common orthogonal basis across multiple domains, proposing two complementary end-to-end differentiable frameworks. First, we introduce Common Principal Component Network (CPCNet), which unrolls the iterative Flury-Gautschi (FG) algorithm into fully differentiable neural layers, enabling covariance-based joint diagonalization within end-to-end representation learning. Second, to address rank deficiency arising from mini-batch covariance estimation in small-sample regimes, we develop Projection Pursuit-based Common Principal Component Network (PP-CPCNet), which directly optimizes a common orthogonal basis without explicit covariance estimation. Experiments validate the joint diagonalization capabilities of both methods, while evaluations on four standard DG benchmarks demonstrate competitive zero-shot generalization performance with favorable computational efficiency.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.12099

SAM-VTP: Video Memory and Measured Geometry for Continuous Traversability Segmentation

Lucas Hoefs, Hui Xie, Senjian An, Fangyi Zhang

原文摘要

Traversability prediction is a core component of autonomous navigation, yet most vision-based methods operate frame-wise and produce temporally inconsistent predictions. We present SAM-VTP (Video Traversability Prediction), which fine-tunes SAM2's video memory pathway for dense, prompt-free traversability segmentation, conditioning each frame on a persistent memory of prior observations. Rather than inferring terrain geometry from appearance, we measure it: slope, roughness and height are computed from depth and fused into the image features. We also introduce an offline continuous supervision pipeline that generates dense per-pixel traversability targets in [0,1] from depth-derived geometry and semantic priors, requiring no robot deployment and no per-pixel annotation, capturing conditional feasibility that binary labels cannot represent. On TartanGround, a large-scale synthetic ground-robot dataset, SAM-VTP improves fuzzy Jaccard from 0.908 to 0.924 and binary IoU from 0.854 to 0.886 over SAM-TP, a frame-wise SAM2 baseline, against a seed-to-seed spread below 0.001, and cuts frame-to-frame flicker by 31%. In closed-loop trials at two field sites, driving the same planner and controller, the baseline's dropouts leave the planner without a feasible path 3.3x more often on a paved course and 6.9x on rough ground, while SAM-VTP runs at 21.4 FPS on the Orin Nano the robot carries.

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2611.01666

Stand-In Effects and Pooling Amplification: Lessons from Biological World Models for Testing Action Effects

Bowen Xu, Boyu Chen, Jiaying Li

原文摘要

Whether an action-conditioned world model responds to actions is usually checked by replacing or removing the actions and measuring the loss in accuracy. Taking biological experiments as the guiding case, we show that such diagnostics report action effects where there are none: under autocorrelated designs, future actions stand in for history the model does not see. When the design law is known, a conditional randomization test written as an e-value is valid in finite samples whatever the model. A second failure arises whenever many sites share one action sequence, as in world models: the pooled score turns small model errors along the action contrast into an e-value near zero, so a valid test loses its power although the effect is real. Tempering the score changes the bet, not the validity. The log e-value is concave in the tempering factor and zero at zero, so too small a factor costs little and too large a one can cost everything; every single factor we tried, estimated or tuned, failed in some setting, and mixtures over a fixed grid never did. Our own preregistered analysis of optogenetic ERK experiments failed by pooling; a tempered statistic, registered before the data were obtained, then rejected the null on a never-opened field (E = 1.1e10). In three registered tests on models trained by others, the pooled test rejected in 0 of 96 units of a diffusion model, and reached the power bar in 10 of 26 Atari games under IRIS's token likelihood and 1 of 6 control tasks under TD-MPC2; mixtures fixed in advance reached it in 96 units, 24 games and 6 tasks. On recorded robot video with iVideoGPT the mixture rejected in 155 of 159 trajectories; the pooled test did not collapse there (143). Under exact nulls the statistics made 14 rejections in 26,664 tests.

cs.LG · q-bio.QM · stat.ME

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.12169

Transactional Data Change for Real-Time Analytics in Lakehouses

Dongyang Geng, Haoqiong Bian, Zinuo Li, Haoyue Li

原文摘要

Real-time data analytics has become a key demand for lakehouses. It requires propagating data changes from the data source to the lakehouse in real time, while ensuring ACID guarantees and minimizing resource consumption and impact on analytical query performance. This is highly challenging for a data system according to the RUM conjecture, especially given the read-optimized immutable file storage layer in lakehouses. Existing lakehouses offer only git-style catalog version control on a single table, and they trade data freshness, query performance, and/or resource efficiency to achieve a barely acceptable data-change throughput. In this paper, we propose a novel data synchronization framework, Retina, that enables real-time, cost-efficient, query-performant, and fully transactional data changes in lakehouses. We implement and open-source Retina. Evaluations show that Retina delivers 10-millisecond-level data freshness and over 3.2 million row/s linear-scalable data-change throughput with low resource consumption and query-performance overhead, significantly outperforming state-of-the-art lakehouse systems and lake-compatible HTAP databases.

cs.DB · cs.AI · cs.DC · cs.PF

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2604.01502

Conformal Risk Control under Non-Monotone Losses: Theory and Finite-Sample Guarantees

Tareq Aldirawi, Yun Li, Wenge Guo

原文摘要

Conformal risk control (CRC) provides distribution-free guarantees for controlling the expected loss at a user-specified level. Existing theory typically assumes that the loss decreases monotonically with a tuning parameter that governs the size of the prediction set. However, this assumption is often violated in practice, where losses may behave non-monotonically due to competing objectives such as coverage and efficiency. In this paper, we study CRC under non-monotone loss functions when the tuning parameter is selected from a finite grid, a setting commonly arising in thresholding and discretized decision rules. Revisiting a known counterexample, we show that the validity of CRC without monotonicity depends critically on the relationship between the calibration sample size and the grid resolution. In particular, reliable risk control can still be achieved when the calibration sample is sufficiently large relative to the grid size. We establish a finite-sample guarantee for bounded losses over a grid of size $m$, showing that the excess risk above the target level $α$ scales on the order of $\sqrt{\log(m)/n}$, where $n$ is the calibration sample size. A matching lower bound demonstrates that this rate is sharp for the CRC selection rule under the same population-safe endpoint condition. We also derive refined guarantees under additional structural conditions, including Lipschitz continuity and monotonicity, and extend the analysis to settings with distribution shift via importance weighting. Numerical experiments on synthetic multilabel classification and real object detection data illustrate the practical implications of non-monotonicity. Methods that explicitly account for finite-sample uncertainty achieve more stable risk control than approaches based on monotonicity transformations, while maintaining competitive prediction set sizes.

stat.ML · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2603.26129

Approximation Algorithms for Scheduling Unrelated Crowd Workers with Contact Delays in Mobile Social Networks

Chi-Yeh Chen

原文摘要

This paper studies a distributed task scheduling problem in which a requester assigns independent tasks to heterogeneous mobile devices (crowd workers) that it reaches only through opportunistic contacts in a mobile social network. The problem generalizes scheduling on unrelated parallel machines: the service time of a task depends on the crowd worker that performs it, every task incurs a worker-dependent contact delay, and the objective is to minimize the total weighted completion time. For identical crowd workers, we tighten the approximation ratio of the Largest-Ratio-First (LRF) algorithm and update the competitive ratio of its online version. For unrelated crowd workers, we show that the time-indexed rounding framework of Im and Li remains valid when worker-dependent contact delays are added, which yields a randomized algorithm with an expected approximation ratio of 1.45; this improves upon the expected ratio of $1.5+ε$ obtained in our previous work for identical crowd workers. We also derandomize this algorithm. To improve computational efficiency, we further propose an algorithm that exploits the optimality of the Smith-ratio order on a single crowd worker, and we prove that it is a deterministic 1.5-approximation algorithm and that its polynomial-time version achieves an approximation ratio arbitrarily close to 1.5. Experiments on synthetic instances and on instances derived from real contact traces show that this algorithm outperforms three variants of LRF in all tested settings, with the largest advantage when the crowd workers are highly heterogeneous, performs on par with a local-search heuristic for unrelated machines, and is within 0.4\% of the optimum on average on small instances that can be solved exactly.

cs.DS

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.12865

Shortcuts and Boundaries: Verifiable RL Misalignment and Contextual Safety Learning

Xinjie Shen

原文摘要

Reinforcement learning (RL) can improve an agent's task capability while reinforcing violations of explicitly stated contextual requirements, even when reward accurately measures useful work. We study this reward-driven misalignment through verifiable reward--context conflicts, where compliant completion is feasible but executable violations earn more reward than every compliant strategy. Encoding these conflicts in task dynamics supports scalable environment construction and compliance feedback from execution histories. In resource and reporting workflows, task optimization increases real productivity through unauthorized resource use or raises reported scores while actual work declines. We observe both amplification of existing violations and the emergence of violations on tasks where none were observed before training. Controlled changes in the training distribution show that minority conflict exposure can still lead to widespread shortcut acquisition, reaching 97.9% certified profitable violations on an API task under a 25% conflict mixture after 128 updates. The same contextual checks supply a simple binary penalty that supports joint capability and compliance gains and partially mitigates acquired violations. Learned safety adapts to expanded authorization and transfers across mechanisms into workflows with evolving evidence and approval requirements. On 32 unseen changing-evidence tasks, resource-trained safety increases safe completion from 53.1% to 89.1% relative to task-only training, while both policies achieve full real work. These results connect a verifiable account of reward-driven misalignment with scalable task construction and safety learning that remains useful under changing execution requirements.

cs.AI · cs.CR

查看论文
arXiv ↗PDF ↗
版本更新 · v5arXiv:2605.12374

Fill the GAP: A Granular Alignment Paradigm for Visual Reasoning in Multimodal Large Language Models

Yanting Miao, Yutao Sun, Dexin Wang, Mengyu Zhou, Pascal Poupart, Lei Lv, Qi Zhao, Li Wang 等 11 位作者

原文摘要

Visual latent reasoning lets a multimodal large language model (MLLM) create intermediate visual evidence as continuous tokens, avoiding external tools or image generators. However, existing methods usually follow an output-as-input latent paradigm and yield unstable gains. We identify a feature-space mismatch behind this instability: dominant visual-latent models build on pre-norm MLLMs and reuse decoder hidden states as predicted latent inputs, even though these states occupy a substantially different norm regime from the input embeddings the model was trained to consume. This mismatch makes direct latent feedback unreliable. Motivated by this diagnosis, we propose GAP, a Granular Alignment Paradigm for visual latent modeling. GAP aligns visual latent reasoning at three levels: data-level alignment grounds latent targets with inspectable auxiliary visual supervision; feature-level alignment maps decoder outputs into input-compatible visual latents through a lightweight PCA-aligned latent head; and model-level alignment assigns latent supervision selectively to examples where the base MLLM struggles. On Qwen2.5-VL 7B, GAP achieves the best aggregate perception and reasoning scores among all compared methods and is the only evaluated latent method that improves both over the base model. Inference-time interventions further show that generated latents carry task-relevant visual signal rather than merely adding token slots. Code is available at https://github.com/Qwen-Applications/GAP.

cs.CV · cs.AI · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.13777

Omni-JEV: Text and Dual-Stream Speech Decisions

Shengzhou Kong

原文摘要

A user utterance does not uniquely determine the next action in full-duplex dialogue: its consequences depend on the assistant's response state and interaction history. Speech activity, utterance completion, and interruption intent provide related evidence, whose implications for a response require an explicit action policy. We formulate this task at collected-segment endpoints and introduce Omni-JEV, which maps text or role-indexed user and assistant-reference speech to probabilities over five interaction actions. An aligned speech-to-speech understanding backbone encodes the acoustic roles separately and shares its candidate readout with text inputs. Bilingual weak supervision supports cross-entropy fine-tuning followed by reinforcement learning for probability-sensitive decisions (RLCD). We compare Gaussian-logit and categorical Choice RLCD from the same supervised initialization under matched training exposure, observing different trade-offs between classification and probability quality. Gaussian Omni-JEV achieves test accuracies of 84.08% for audio and 96.55% for text. This common action interface supports decisions about when a full-duplex voice assistant should respond, preserve ongoing speech, interrupt, or resume an interrupted response.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2603.05868

AnyCamVLA: Zero-Shot Camera Adaptation for Viewpoint Robust Vision-Language-Action Models

Hyeongjun Heo, Seungyeon Woo, Sang Min Kim, Junho Kim, Junho Lee, Yonghyeon Lee, Young Min Kim

原文摘要

Despite remarkable progress in Vision-Language-Action models (VLAs) for robot manipulation, these large pre-trained models require fine-tuning to be deployed in specific environments. These fine-tuned models are highly sensitive to camera viewpoint changes that frequently occur in unstructured environments. In this paper, we propose a zero-shot camera adaptation framework without additional demonstration data, policy fine-tuning, or architectural modification. Our key idea is to virtually adjust test-time camera observations to match the training camera configuration in real-time. For that, we use a recent feed-forward novel view synthesis model which outputs high-quality target view images, handling both extrinsic and intrinsic parameters. This plug-and-play approach preserves the pre-trained capabilities of VLAs and applies to any RGB-based policy. Through extensive experiments on the LIBERO benchmark, our method consistently outperforms baselines that use data augmentation for policy fine-tuning or additional 3D-aware features for visual input. We further validate that our approach constantly enhances viewpoint robustness in real-world robotic manipulation scenarios, including settings with varying camera extrinsics, intrinsics, and freely moving handheld cameras. Project Page: https://heo0224.github.io/AnyCamVLA/

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.04726

Bridging Modalities and Tasks: A Unified Hierarchical ViT for SAR-to-Optical Translation and Semantic Segmentation

Siyuan Liu, Xuze Zhang, Yongshun Wang, Licong Pan, Hang Liu, Huihui Li

原文摘要

Synthetic Aperture Radar (SAR) remains available through cloud and darkness, yet speckle and non-intuitive scattering make it hard to inspect and interpret. SAR-to-optical (S2O) translation renders radar observations as optical-like images for visual access, while semantic segmentation on SAR remains necessary since land-cover decisions cannot rely on synthesised pixels. We propose BMT, a hierarchical Vision Transformer coupling both tasks on a shared encoder via multi-scale one-hot conditioning, a gated LocalViTBlock, a colour-calibrating generation head, ControlNet-style zero-initialized injection, and uncertainty-based task weighting. The joint framework's value is not label-free training: it uses paired optical images and the same pixel annotations as a dedicated SAR segmentor, supervising the SAR decoder and teacher-forcing the generator with ground-truth one-hot maps. At test time, the model takes only SAR and returns an optical-like readout and a radar-decided land-cover map, with the translation objective supplying dense appearance supervision that speckle lacks. On WHU-OPT-SAR, the predicted-condition model achieves the best SSIM and PSNR among compared translators; its land-cover map surpasses the same backbone trained for labelling alone and is within about one point of DeepLabV3+.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.10238

Geometry-Supervised Visual Representation Learning for Multi-Phenotype Lesion Interpretation in Medical VLMs

Hao Wang, Qiwei Zeng, Shuchang Ye, Jinghao Lin, Yuezhe Yang, Yige Peng, Jinman Kim, Lei Bi

原文摘要

Medical vision-language models (VLMs) have shown increasing potential for clinical image interpretation. However, these models still struggle to interpret multi-phenotype lesions whose diagnosis requires the joint assessment of multiple pathological phenotypes. Existing vision-language alignment methods produce visual representations that fail to preserve anatomical hierarchies and relationships among phenotypic subclasses. This stems from their reliance on semantic supervision, which lacks geometric constraints to preserve these relationships in the visual embedding space. Moreover, the sparsity of lesion-related anatomical and phenotypic representations makes it difficult for medical VLMs to capture important diagnostic evidence. To address these limitations, we propose \textbf{PureVision}, a geometry-supervised visual representation learning framework for multi-phenotype lesion interpretation in medical VLMs. It combines a geometry-supervised representation learning module, \textbf{PureEyes}, and an anatomy-guided evidence aggregation module, \textbf{PureNeurons}. PureEyes provides geometric supervision through ideal spatial distributions that encode anatomical hierarchies and phenotypic subclass relationships. PureNeurons projects visual representations into the learned latent space, using their positions to selectively aggregate lesion-specific anatomical and phenotypic evidence. Experiments on \textit{LIDC-IDRI}, \textit{CBIS-DDSM}, and \textit{3DReasonKnee} demonstrate that PureVision improves lesion grounding and phenotype characterization in visual question answering and radiology report generation. Code is available at: https://anonymous.4open.science/r/purevision-06C2.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.24206

SWE-Bench Pro V2: A Cleaner, Harder-to-Game Coding Benchmark

Soham Dan, Jeff Da, Sanyam Bhutani, Miguel Romero Calvo, Ivan Guevara, Fernando Carabedo, Jose Bernal, Julio Reyna 等 11 位作者

原文摘要

SWE-Bench Pro is a widely used benchmark of long-horizon software engineering tasks for coding agents. As agents improve, they reach task issues earlier agents never touched and interact with the evaluation environment in ways the original harness did not anticipate. We present SWE-Bench Pro V2, a revised public split built to keep the benchmark accurate as capabilities advance. Agentic audits, expert adjudication and blind re-implementation repaired or retired 480 of the 731 tasks: 23 contracted engineers spent 404 hours, each working in a language they write professionally, and every revised instruction was verified by a second engineer solving the task from the text alone. A hardened protocol restricts network access to the model endpoint and re-grades every patch in a clean container, so agents cannot retrieve upstream fixes or get credit for changes outside their patch. On the 642 V2 tasks frontier agents resolve up to 99.4% while other systems stay clearly separated, and a 51-task Hard split spreads systems from 52.9% to 98.0%. We release the benchmark and distill lessons for maintaining agent benchmarks over time.

cs.SE · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2509.21284

Bounds of Chain-of-Thought Robustness: Reasoning Steps, Embed Norms, and Beyond

Dingzirui Wang, Xuanliang Zhang, Keyan Xu, Qingfu Zhu, Wanxiang Che, Yang Deng

原文摘要

Existing research indicates that the output of Chain-of-Thought (CoT) is significantly affected by input perturbations. Although many methods aim to mitigate such impact by optimizing prompts, a theoretical explanation of how these perturbations influence CoT outputs remains an open area of research. This gap limits our in-depth understanding of how input perturbations propagate during the reasoning process and hinders further improvements in prompt optimization methods. Therefore, in this paper, we theoretically analyze the effect of input perturbations on the fluctuation of CoT outputs. We first derive an upper bound for input perturbations under the condition that the output fluctuation is within an acceptable range, based on which we prove that: (i) This upper bound is positively correlated with the number of reasoning steps in the CoT; (ii) Even an infinitely long reasoning process cannot eliminate the impact of input perturbations. We then apply these conclusions to the Linear Self-Attention (LSA) model, which can be viewed as a simplified version of the Transformer. For the LSA model, we prove that the upper bound for input perturbation is negatively correlated with the norms of the input embedding and hidden state vectors. To validate this theoretical analysis, we conduct experiments on three mainstream datasets and four mainstream models. The experimental results align with our theoretical analysis, empirically demonstrating the correctness of our findings.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2602.08294

Not All Gap Correction Helps: A Geometric View of External Information in LLM Inference

Dingzirui Wang, Xuanliang Zhang, Keyan Xu, Qingfu Zhu, Wanxiang Che, Yang Deng

原文摘要

Providing external information beyond the query is a common way to improve large language model (LLM) inference, as seen in in-context learning (ICL), retrieval-augmented generation (RAG), and memory-based methods (Mem). Existing work increasingly suggests that useful external information should compensate for the gap between LLMs and user queries, implicitly assuming that stronger gap-correction can lead to better performance. In this paper, we challenge this assumption and show that more gap-corrective external information can instead harm performance. We analyze this phenomenon in Transformers through the lens of reasoning error, defined as the difference between the predicted answer vector and the ground-truth answer vector. Our analysis reveals that external information can be not only under-corrective but also over-corrective, where excessively strong correction increases reasoning error. We further show that the induced correction vector is jointly determined by attention and the complementarity between the external information and the query, and derive conditions under which information is properly corrective in both direction and magnitude. Experiments on four mainstream LLMs and seven reasoning benchmarks across ICL, RAG, and Mem validate our theory and support a theory-guided information selection method.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.21836

Modeling and Force Analysis of a Soft Linear Pneumatic Actuator

Mohammed Abboodi

原文摘要

Soft sleeve actuators (SSAs) have recently been developed as a pneumatic actuation approach for wearable and assistive robotic systems. By integrating the actuation structure into a sleeve-like geometry, these actuators can reduce reliance on external attachment layers and transmission mechanisms while maintaining compliance with limb-shaped surfaces. However, the force-generation behavior of SSAs remains insufficiently explained, particularly with respect to the variation of output force during extension, the influence of external loading, and the mechanical role of axial stiffness. This paper presents an analytical and experimental force analysis of a linear soft sleeve actuator (LSSA). A quasi-static analytical model was developed by expressing the net axial force as the pressure-generated contribution from the cap and folded walls, reduced by the force associated with axial stiffness. The model incorporates internal pressure, projected pressure areas, folded wall geometry, axial displacement, and an experimentally fitted axial stiffness relation. Prescribed-extension and static-load experiments were conducted to evaluate the actuator response. At 125 kPa, the generated force decreased from approximately 112 N at zero extension to nearly zero at 40 mm. Static loading delayed measurable force generation and reduced force output, particularly at low and intermediate pressures. The results show that LSSA force generation is governed by coupled effects of pressure, geometry, displacement, loading, and axial stiffness.

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.25109

Soft Pneumatic Actuators for Robotics: From Motion Generation to Performance Optimization

Mohammed Abboodi

原文摘要

Soft pneumatic actuators are widely used in soft robotics because they can produce large motions while remaining compliant enough to interact safely with objects, environments, and the human body. However, their performance is not solely determined by pressure. Instead, the response depends on the way the actuator is built, including the shape of its chambers, the placement of reinforcements, the use of folds, material stiffness, and the constraints that guide its deformation. As the literature has expanded, it has become more difficult to determine which mechanism is most suitable for a given application and which reported results can be compared across studies. This review examines soft pneumatic actuators according to the design strategies used to generate four motion classes: linear, bending, twisting, and omnidirectional actuation. For each class, it analyzes the structural features that define the deformation path, including braid angle, fold geometry, fiber orientation, chamber arrangement, structural asymmetry, and internal constraint layers. It then discusses how the design choice affect motion output, force generation, air demand, repeatability, durability, fabrication difficulty, and robotic integration. The review further identifies key conditions that must be considered when selecting or comparing actuators, including pressure, loading condition, actuator size, pneumatic supply, and hysteresis This approach helps explain why actuators with similar motion outputs may differ substantially in design requirements, pneumatic demand, and practical suitability. It also highlights the design priorities needed for compact, efficient, repeatable, and deployable soft pneumatic systems in wearable, biomedical, and mobile robotic applications.

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v6arXiv:2610.03228

Reference-Compatible Parallel Execution for Quantized LLM Inference

Chengzhan Li

原文摘要

Batching a quantized large language model (LLM) can change full-vocabulary outputs and the key-value (KV) cache even when token positions and causal masks are unchanged. We introduce reference-derived phase plans: per-call execution descriptions that preserve a fixed deployment's producer, physical attention range, arithmetic and state ownership while exposing parallel work. Whole-graph plans split the complete computation at compatibility boundaries; local-attention plans split only attention and retain wider projection and feed-forward work. Independent arithmetic groups can also be packed into one operation without changing each group's reduction. We implement and qualify these paths on CUDA and Vulkan for Qwen, Llama and SmolLM2. On Linux/RTX A5000, compatible speculative execution reduces resident-service time by 17.21 percent, 25.54 percent and 11.52 percent relative to the fastest measured fixed target-only control for Qwen3-8B, Qwen3-14B and Llama-8B. Balanced experiments isolate an increment of 0.80 to 1.64 percent from Q8 projection packing and smaller, geometry-dependent gains from local cuts. For Qwen3-32B with mixed Q4_K/Q6_K weights, independent-channel grouping reduces ordinary service time by 5.61 percent against a graph-level scalar-column loop. Equal-work tests favor eight-query execution in aggregate, whereas complete live requests favor shorter proposals because longer proposals increase target work. The results show how to exploit known-query parallelism within an established numerical reference and how complete-request costs determine the useful execution strategy.

cs.PF

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.09074

TAP: Efficient Long-Horizon Agent Pruning via Trajectory-Anchored Recovery

Yuanzhe Li, Pengxin Wang, Yuxin Ren, Jianing Deng, Jingtong Hu, Song Wang, Jingdi Chen, Huanrui Yang

原文摘要

Emerging long-horizon agentic tasks require repeated model calls, worsening the inference cost of already-costly language models. While narrow agentic tasks suggest potential for aggressive model pruning without performance drop, empirical results show existing methods proposed for question answering tasks severely degrade task performance when applied to agentic models. We trace this failure to two decisions: what to prune and how to recover. For pruning, one-shot importance estimates fail to track how the pruned model adapts. For recovery, offline distillation covers only teacher prefixes, while full-trajectory on-policy distillation causes student errors to compound across turns. In this work, we propose Trajectory-Anchored Pruning (TAP), the first structural pruning framework for reinforcement learning (RL)-trained agents. TAP couples structural pruning with efficient on-policy recovery, anchoring interactions to teacher trajectories while allowing the student to generate each reasoning-action response. A frozen dense teacher supervises the student's response prefixes, addressing within-response training-inference mismatch while preventing student-induced deviations from propagating across training turns. Instead of one-shot pruning, TAP re-scores channels using gradients of the recovery objective on the recovered student, connecting iterative channel selection to the evolving policy. With 60% of FFN channels removed, TAP retains 99.2% and 88.0% of the dense 7B agents' task success rates on ALFWorld and WebShop, respectively, while reducing GPU time per successful task by approximately 22% and 17%. These results demonstrate effective structural compression of long-horizon agents under a limited recovery budget.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2611.02433

Blueprint2CAD: Benchmarking Foundation Models on Engineering Drawing-to-CAD Render Correspondence

Alexander Valverde, Danny Xie-Li, Zhichao Wang, Marcos V. Conde

原文摘要

Engineering drawings specify mechanical parts through orthographic projections. Recognizing which part a drawing depicts underlies part search, design reuse, and digitization, yet no public benchmark measures whether vision models can do it. We introduce Blueprint2CAD, a benchmark for drawing-to-render correspondence: given an orthographic line drawing and a shaded render, usually from another viewpoint, decide whether they show the same part. Its data layer, BlueCAD-30K, applies an open drafting pipeline to 29,925 Zero-to-CAD models, adding view-aligned renders, clean and dimensioned drawings from six viewpoints, and surface-type metadata. Frozen encoders retrieve the matching part from a 10,000-model gallery with up to 65% Recall@1 when viewpoints are known but 43% when they are not, and the four encoders we test lie within six points of each other: viewpoint, not backbone, is the bottleneck. Six zero-shot vision-language models reach only 0.52-0.67 verification accuracy, with model-specific Yes/No biases that chain-of-thought shifts but does not remove; a quick LoRA fine-tuning on in-domain pairs improves verification accuracy by a few points but sharpens these biases and leaves the viewpoint failure intact.

cs.CV

查看论文
arXiv ↗PDF ↗