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

+1085新增论文
+884其中 AI 相关
+130已有论文修订
+94新增 AI 导读

arXiv 最近完成扫描:2026-11-19 06:59 · 全站最近入库:2026-11-19 06:59各信源进度 →

统计口径与数据新鲜度

统计 2026-11-18 11:24 — 2026-11-19 11:24(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:2026-11-19 06:59最近发现新增版本:2026-11-19 06:59库内论文最新更新时间:2026-11-18 08:00

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。

26,871 已收录论文22,375 AI 相关8,156 有修订112 HF 社区发现

论文修订列表

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 8,156 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-19 06:57(北京时间)
版本更新 · v3arXiv:2602.01932

Things that Matter -- Identifying Interactions and IoT Device Types in Encrypted Matter Traffic

Kristopher Alex Schlett, Bela Genge, Ioan Padurean, Savio Sciancalepore

原文摘要

Matter is the latest interoperability standard for the consumer Internet of Things (IoT) market, developed by the Connectivity Standards Alliance (CSA). Matter's design pays particular attention to security and privacy: it provides validated secure session establishment protocols, and it uses robust security algorithms to encrypt communications between IoT devices and Matter controllers. However, little is known about the extent to which information can still be inferred from encrypted traffic. This paper investigates whether a passive adversary can identify Matter interactions and device types by analyzing encrypted traffic metadata. Using datasets collected from both real-world and simulated Matter deployments, we identify repeatable fingerprints based on packet lengths, directions, and packet sequences. Our results show that Matter interactions can be identified with over 95% accuracy even under packet loss and delay conditions. Furthermore, device types such as lighting, lock, and plug appliances can be inferred with accuracy exceeding 88%. These findings reveal a significant privacy risk affecting Matter-enabled deployments regardless of vendor. The CSA acknowledged our findings and indicated that mitigation mechanisms may be considered in future Matter versions.

cs.CR

版本更新 · v2arXiv:2603.16568

Manifold-Matching Autoencoders: Embedding Data into Chosen Geometric Structures

Laurent Cheret, Vincent Létourneau, Isar Nejadgholi, Chris Drummond, Hussein Al Osman, Maia Fraser

原文摘要

Manipulating the structure of data in the latent space of an autoencoder is a long-standing goal of representation learning. We propose a simple regularization, Manifold Matching, that imposes an explicit geometric structure on that latent space. The idea is to take a reference representation R of the data, which need not be the raw input X fed to the network, and to minimize, batch-wise, the mean squared error between the pairwise distances in R and those in the latent space Z. Although the autoencoder still learns to reconstruct X, the shape of its latent space comes to mimic that of R. The objective requires no labels, only reference distances, and because those are scalars, the dimensionality of R is decoupled from that of Z: any reference can guide any bottleneck. Using this single instrument we study how different geometric structures can be imposed on a latent space, comparing against autoencoders built on geometric or topological regularizers. Across fifteen datasets under a shared protocol with statistical testing, the resulting Manifold-Matching Autoencoder (MMAE) matches the global-structure quality of persistence-based autoencoders, statistically tying RTD-AE on distance correlation and triplet accuracy, and stays competitive on the local and topological metrics that other methods lead, at the computational cost of a plain autoencoder: it uses only a batched distance loss, with no persistence computation. We frame MMAE not as a universal winner but as an efficient and controllable regularizer whose reference selects which structure the latent inherits, and which turns any fixed embedding into a reusable out-of-sample encoder. All experiments use a two-dimensional bottleneck.

cs.LG · cs.AI

版本更新 · v2arXiv:2610.17590

Revisiting the 16x Pretraining Compute Gap between Diffusion and Autoregressive Language Models

Guanghan Wang, Bing He, Zhan Shi, Benoit Dumoulin, Hanqing Lu, Volodymyr Kuleshov

原文摘要

Diffusion language models (DLMs) have attracted wide attention for their fast, parallel decoding, but they are believed to be far less efficient to train than autoregressive language models (ARMs): a widely cited study reports that a DLM needs roughly 16x the training compute of an ARM to match its likelihood, a figure that has made pretraining DLMs from scratch look prohibitive at scale. We find that the methodology behind this figure has fallen out of step with current DLM practice: it scores the model with its evidence lower bound (ELBO) and trains a full-sequence, decoder-only model. We therefore revisit the scaling laws of DLMs, adopting the exact likelihood under the model's own decoding order, block diffusion training, an encoder-decoder architecture, and a sweep over the block sizes used in practice. Under this new protocol, we study across ten model variants, about 1,700 models in total, and find that pretraining DLMs is far cheaper than previously thought. Concretely, scoring with the exact likelihood reduces the compute multiple to 12.8x; block diffusion training and an encoder-decoder architecture further reduce it to 7.8x at block size 32, and smaller block sizes narrow it to 4.6x at block size 8 and 3.0x at block size 4. We complement these scaling analyses with an empirical rule for the optimal decoder depth of encoder-decoder block diffusion, a held-out compute budget beyond the fitting range at which the gap between variants persists, and downstream benchmark results where likelihood improvements carry over to accuracy. Overall, the 16x figure substantially overstates the pretraining cost gap between DLMs and ARMs, and pretraining DLMs from scratch is far more viable than previously assumed.

cs.LG · cs.AI

版本更新 · v5arXiv:2604.01650

AromaGen: Interactive Generation of Rich Olfactory Experiences with Multimodal Language Models

Awu Chen, Silang Wang, Yunge Wen, Jianing Yu, Zinab Eisa, Ana Julia Santos Lopes, Jas Brooks, Hiroshi Ishii 等 9 位作者

原文摘要

Smell's connections with food, memory, and social experience have motivated researchers to bring olfaction into interactive systems. Multimodal AI opens new possibilities for generating smells from natural language, yet it remains unclear whether pretrained models encode sufficient olfactory knowledge to translate language into perceptible compositions. We present AromaGen, an AI-powered 12-channel wearable olfactory system that maps free-form natural-language descriptions to 12 base odorants selected to cover a semantically derived olfactory space, while supporting iterative refinement through natural-language feedback. In a between-subjects study (N=60) across 50 real-world benchmark smells, participants distinguished AromaGen-generated target smells above chance in a three-alternative forced-choice task, with retrieval-augmented AI composition performing comparably to human-expert composition and zero-shot AI composition. Natural-language feedback significantly improved the perceived similarity of both AI- and human-expert compositions. Our findings demonstrate the feasibility of using pretrained multimodal AI, with human aroma composition data, to generate perceptible olfactory experiences from natural language.

cs.HC · cs.AI

版本更新 · v2arXiv:2606.09258

Back to the Familiar Future: Failure Recovery for VLA Policies via Pre-Imagined Milestone Selection

Suyeon Shin, Juwon Kim, Hyeonbin Park, Hyunseo Kim, Hyundo Lee, Hyung-Sin Kim, Byoung-Tak Zhang

原文摘要

Vision-language-action (VLA) policies can deviate from nominal trajectories during manipulation, even when tasks remain physically feasible. Recovering from these deviations is challenging, as they push the policy into unfamiliar state spaces where direct re-planning frequently destabilizes action sequences. We propose Back to the Familiar Future (B2FF), a recovery framework for foresight-driven VLAs that leverages future visual conditioning as a recovery interface. Before execution, the VLA generates a milestone bank of familiar future states conditioned on the clean initial observation. At recovery time, a recoverability-aware selector selects a recovery milestone from this bank and enforces it as a fixed visual goal. This enables the VLA to robustly map off-trajectory observations back to a familiar future. On failure-injected LIBERO, under controlled recovery timing aligned with the injected failure, B2FF increases the average success rate of a baseline VLA from 56.3% to 74.0%, demonstrating that pre-imagined milestones can guide recovery without fine-tuning the low-level action generator.

cs.RO

版本更新 · v2arXiv:2610.32503

Pareto-optimal Non-Clairvoyant Scheduling with Predictions

Ziyad Benomar, Vianney Perchet

原文摘要

We study preemptive single-machine non-clairvoyant scheduling with predictions, where a prediction of the order of the jobs (or of their sizes) is available, and the objective is to minimize the total completion time. An algorithm is $c$-consistent if it is $c$-competitive when the prediction is correct, and $R$-robust if it is $R$-competitive whatever the prediction. Since the introduction of this problem as one of the first examples of learning-augmented algorithms, the best possible trade-off between consistency and robustness has remained open. We determine it exactly for deterministic algorithms. For every $R>2$, we give an $R$-robust algorithm whose consistency is \[ c(R)=R\Big(1-\Big(1-\frac{R+2}{R^2}\Big)^{\frac R{R+2}}\Big), \] and we prove that no deterministic $R$-robust algorithm, with either predicted order or predicted job sizes, has better consistency. In particular, consistency $1+\varepsilon$ is achievable with robustness $\varepsilon^{-1/2}+\frac16+O(\sqrt\varepsilon)$, whereas all previous algorithms required robustness $Θ(1/\varepsilon)$. Our algorithm is greedy: it follows the predicted order as fast as possible, subject to the invariant that the cost paid so far is at most $R$ times the optimum of the instance in which every unfinished job would end immediately. The matching lower bound rests on the observation that every robust deterministic algorithm must satisfy the same invariant along its run on correct predictions. Furthermore, we give a simple randomized algorithm, using predicted sizes, with consistency $1+\varepsilon$ and robustness $(3\varepsilon)^{-1/2}+O(1)$, strictly below the deterministic frontier. The question of determining the Pareto-optimal frontier among randomized algorithms remains open.

cs.DS · cs.GT

版本更新 · v5arXiv:2403.15191

Your Trust, Your Terms: A General Paradigm for Near-Instant Cross-Chain Transfer

Di Wu, Jingyu Liu, Xuechao Wang, Jian Liu, Yingjie Xue, Kui Ren, Chun Chen

原文摘要

Cross-chain transactions today remain slow, costly, and fragmented. Existing custodial exchanges expose users to counterparty and centralization risks, while non-custodial liquidity bridges suffer from capital inefficiency and slow settlement; critically, neither approach guarantees users a unilateral path to recover assets if the infrastructure fails. We introduce the Delegated Ownership Transfer (DOT) paradigm, which decouples key ownership from value ownership to enable secure, high-performance cross-chain payments. In DOT, a user deposits funds into a sandboxed on-chain Temporary Account (TA) (value ownership) while delegating its private key (key ownership) to an abstract Trusted Entity (TE). Payments and swaps are thus reframed as near-instant, off-chain ownership handoffs. Security follows from dual guarantees: the TE's exclusive control prevents double-spending, while a pre-signed, unilateral recovery transaction ensures users retain ultimate authority over their assets. Building on this foundation, we design a novel off-chain atomic swap that executes optimistically in near real time and remains fair even if the TE fails. We formalize the security of DOT in the Universal Composability framework and present two concrete instantiations: a high-performance design based on Trusted Execution Environments (TEEs) and a cryptographically robust variant leveraging threshold cryptography. Our geo-distributed prototype shows that cross-chain payments complete in under 16.70 ms and atomic swaps in under 33.09 ms, with costs fully decoupled from Layer-1 gas fees. These results provide a practical blueprint for building secure, efficient, and interoperable cross-chain payment systems.

cs.CR · cs.DC

版本更新 · v2arXiv:2606.18439

RegimeVGGT: Layer-Wise Spatially Preserving Redundancy Removal for Visual Geometry Grounded Transformer

Jinhao You, Shuo Lyu, Zhuohang Lyu, Tanxuan Li, Zibo Zhao, Jiaxiang Hu, Kai Tang, Yichen Guo

原文摘要

Visual Geometry Grounded Transformer (VGGT) recovers dense 3D scene structure from multi-view images in one forward pass, but quadratic cross-frame attention limits its scalability. Existing training-free accelerators exploit token, frame, or attention redundancy; we investigate how layer-wise functional roles can guide compression. Our spectral, probing, and causal analyses reveal three regimes: shallow layers lack cross-view structure, middle layers drive cross-view alignment, and deep layers are redundant for dense geometry yet their cross-frame attention remains essential for pose. RegimeVGGT applies layer-wise U-shaped compression along two axes: Saliency-Guided Banded Merging protects geometry- and edge-salient tokens, while Selectively Protected K/V Downsampling preserves cross-frame spatial coverage and the pose-critical path through a phase-shifted spatial grid, a reference-frame anchor, and uncompressed camera/register tokens. Without any training, RegimeVGGT achieves a 6.7x speedup over the memory-optimized VGGT* at matched reconstruction quality. Project page: https://tracyou07.github.io/regimevggt/

cs.CV · cs.RO

版本更新 · v2arXiv:2605.23135

The Impact of AI Coding Assistants on Software Engineering: A Longitudinal Study

Annie Vella, Kelly Blincoe

原文摘要

AI coding assistants have become prolific in recent years. Through a longitudinal mixed-methods investigation, we examined how professional software engineers perceive the effects of AI coding assistants in regard to task focus, developer experience, and productivity. Two questionnaires were administered six months apart, yielding 158 eligible participants at the first time point, 101 at the second, and a matched longitudinal cohort of 95. Participants reported spending less time on most development tasks, with 82% reporting less on writing code. We found a broader shift in focus from creation to verification activities. We propose a new category of work we term supervisory engineering work, encompassing the direction, evaluation, and correction of AI output. This construct is broader than the extra review overhead that has been previously reported in literature since it encompasses directing and shaping AI output both before and during creation. We also identified a productivity-experience paradox: consistently positive productivity perceptions alongside eroding developer experience. These findings suggest that AI coding assistants may be reshaping both the nature of software engineering work and how engineers experience it.

cs.SE

版本更新 · v2arXiv:2610.37652

基于学习安全滤波器的灵巧手内物体位姿到达

Dexterous In-Hand Object Pose Reaching with Learned Safety Filters

Ian Chuang, Hiya Shah, Vamsi Chilamkur, Iman Soltani, Claire Tomlin

AI 速读 基于摘要

面向LEAP灵巧手的手内物体位姿到达任务,作者提出无需人类数据的仿真训练方案,使策略零样本迁移到真实手,并叠加可调保守度的学习残差安全滤波器以减少掉落。

cs.RO · cs.LG · eess.SY

查看原文摘要

Dexterous manipulation with multi-fingered hands is the next frontier of robotics, yet learned hand skills remain limited to simple grasps or basic repeated motions, such as rotating an object. As robots attempt more demanding in-hand maneuvers, safety, such as not dropping the object, becomes critical, yet it is rarely measured. We study in-hand object pose reaching, a task that demands diverse and difficult grasps: moving an object to almost any position and orientation within reach of the fingers, at any wrist orientation, through long sequences of consecutive goals. We present a recipe for training such policies for a LEAP hand in simulation without human data, built on a range of insights, from diverse resets and goals via generated grasp datasets to an adaptive friction curriculum. With careful system identification, the policies transfer zero-shot to the real hand, using either motion capture or a single wrist-mounted depth camera, and prove remarkably capable and robust. On top of the task policy, we add a learned residual safety filter with adjustable conservativeness, letting us tune the balance between performance and safety and greatly reduce object drops. In 13.3 hours of real-robot evaluation over 4,800 unseen goals across 4 objects and diverse wrist orientations, the state-based policy reaches 95.2% of goals within 2 cm and 0.4 rad, and the filter cuts its drops by 75% while still reaching 91.5% of goals. The policies can also spin a lightbulb held in an arbitrary pose, reach out-of-distribution goals, withstand fast arm motion and perturbations, and, needing only a wrist camera, serve as portable skills such as screwing in a lightbulb. We release the full recipe as a platform for studying performance, reliability, and safety in high-dimensional dexterous systems. Videos and more are available at https://ian-chuang.github.io/dexsafe/.

版本更新 · v2arXiv:2607.10682

SensorPerch:让感知随时随地发生

SensorPerch: Sense Wherever and Whenever it Matters

Zhanxin Wu, Ruofei Tong, Tapomayukh Bhattacharjee

AI 速读 基于摘要

提出将传感器视为可被机器人自主拆卸与重新附着于环境的独立物理实体,并实现轻量无线可重构传感器平台与视角选择框架,以按需构建任务相关视角。

cs.RO

查看原文摘要

Existing robotic perception is constrained by sensors that are either robot-mounted or permanently fixed in the environment, locking perception to a limited set of viewpoints. Yet as robots perform increasingly diverse tasks, the most informative viewpoint shifts from one task to the next-often somewhere onboard sensor and static infrastructure can not readily satisfy. To address this gap, we propose SensorPerch, a novel realization of active perception that decouples sensing from both the robot embodiment and the environment by treating sensors as independent physical entities that the robot can autonomously detach and re-attach within the environment. SensorPerch presents one realization of this paradigm: a lightweight, wireless, reconfigurable sensor platform that can perch on diverse surfaces, paired with a viewpoint-selection framework that determines task-optimal sensor placements. Together, these enable robots to construct task-relevant viewpoints on demand, independent of the robot's current position and available fixed infrastructure. We demonstrate the paradigm on two task classes: (i) object-coupled perception, where SensorPerch enables persistent object-state detection beyond the robot's current position, achieving successful event detection even when the robot is not nearby; and (ii) policy-coupled perception, where SensorPerch allows robots to construct diverse, policy-specific viewpoints for various policies, achieving success rates comparable to those obtained using oracle viewpoints.

版本更新 · v2arXiv:2608.19490

基于自示范生成控制的多任务操作VLA微调

Fine-Tuning VLAs with Self-Demonstrated Generative Control for Multi-Task Manipulation

Prachi Garg, Steve Xing, Prahit Yaugand, Saurabh Gupta, Derek Hoiem

AI 速读 基于摘要

提出一种自监督方法,用零样本VLA生成的在线交互轨迹作为额外训练数据来微调VLA,从而在新机器人上获得多任务操作策略。

cs.RO · cs.CV · cs.LG

查看原文摘要

State-of-the-art vision-language-action (VLA) models such as pi0.5 exhibit strong semantic understanding, instruction following and task behavior. However, when deployed on new robots, even minor mismatches in hardware configuration relative to pretraining can cause severe performance drops. Finetuning the VLA on in-domain expert data from the new embodiment improves performance on the expert task but leads to a loss in its original instruction following and behavioral priors. In this paper, we propose a self-supervised method that generates online interaction rollouts from the zero-shot VLA as additional training data for finetuning. Our experiments show this yields strong multi-task manipulation policies on the target robot that (1) inherit pretrained tasks distilled from the zero-shot VLA without expert demos for them, (2) improve success and generalist instruction following beyond expert-only fine-tuning on expert-teleoperated tasks, and (3) mitigate forgetting of held-out tasks from the pretrained policy, not part of fine-tuning. We demonstrate the success of our approach across test sets probing generalization on a real ALOHA robot and a new simulation benchmark in RoboTwin. Video results are available at https://self-supervised-control.pages.dev/

版本更新 · v2arXiv:2605.09995

语义标注缓解后训练模式坍缩

Annotations Mitigate Post-Training Mode Collapse

Jacob Mitchell Springer, Madhu Advani, Lukas Aichberger, Arwen Bradley, Eran Malach, Omid Saremi, Sinead Williamson, Preetum Nakkiran 等 10 位作者

AI 速读 基于摘要

作者提出 annotation-anchored training,通过在预训练与后训练中保留语义标注分布,缓解监督微调导致的语义模式坍缩并提升多样性。

cs.CL

查看原文摘要

Post-training (via supervised fine-tuning) improves instruction-following, but often induces semantic mode collapse by biasing models toward low-entropy fine-tuning data at the expense of the high-entropy pretraining distribution. Crucially, we find this trade-off worsens with scale. To close this semantic diversity gap, we propose annotation-anchored training, a principled method that enables models to adopt the preference-following behaviors of post-training without sacrificing the inherent diversity of pretraining. Our approach is simple: we pretrain on documents paired with semantic annotations, inducing a rich annotation distribution that reflects the full breadth of pretraining data, and we preserve this distribution during post-training. This lets us sample diverse annotations at inference time and use them as anchors to guide generation, effectively transferring pretraining's semantic richness into post-trained models. We find that models trained with annotation-anchored training can attain $6 \times$ less diversity collapse than models trained with SFT, and improve with scale.

版本更新 · v2arXiv:2605.31245

迈向可识别的稀疏自编码器

Toward Identifiable Sparse Autoencoders

Walter Nelson, Theofanis Karaletsos, Francesco Locatello

AI 速读 基于摘要

作者从理论和实验角度指出稀疏自编码器(SAE)训练不稳定,并通过最小化架构与训练流程改动提出可识别的 iSAE 变体。

cs.LG

查看原文摘要

Recently, sparse autoencoders (SAEs) have emerged as an attractive tool for interpreting and interacting with representations in practical neural networks. While it is common empirical folklore, we also show theoretically that SAEs are highly unstable: different training runs are likely to produce different concept dictionaries and sparse codes. We characterize the model properties that hinder the stability of real-world SAEs, and address each of these problems through minimal changes to the architecture and training procedure. Together, these changes yield two versions of an \textbf{i}dentifiable SAE (iSAE), a variant of the standard TopK SAE with lower reconstruction error and improved stability. We explain this improvement theoretically by connecting SAEs with traditional dictionary learning approaches, and show that the dictionaries learned in practice satisfy an approximate restricted isometry condition, rendering the corresponding sparse codes in those models near-identifiable.

版本更新 · v2arXiv:2606.31004

Where the Real Clips Came From: Corpus Identity in AI-Generated Video Detection

Mert Onur Cakiroglu, Zhihe Lu, Mehmet Dalkilic, Hasan Kurban

原文摘要

Benchmarks for detecting AI-generated video take their real clips from a few curated collections, so a clip's label is collinear with where it came from. We call a score's dependence on this origin corpus identity, and we measure it on four benchmarks and six real collections. Three probes read a file's declared frame types, the motion vectors its codec stores, and the frames' appearance. Frame-type counts alone separate real from generated at AUC 0.999 on GenVidBench, and re-encoding every clip with one setting at a fixed length still leaves 0.61 to 0.63 on three benchmarks. We introduce a fixed frame-type pattern that removes this leak from every clip of at least 2 s. The same re-encode lowers the ReStraV detector on GenVidBench from 0.85 to 0.60, just above the 0.59 its features reach between two real collections, because its 8 fps repeats frames in real clips. Motion vectors alone separate real collections from one another at AUC 0.51 to 0.98, and only one benchmark ships the two labeled real sources needed to measure this separation within it. In deployment, with each threshold set to flag 5% of its training benchmark's real clips, the motion-vector and appearance probes and linear readouts on ReStraV and WaveRep features each flag 15% or more of another real collection. Choosing the threshold on real videos from the target collection brings the rate back to about 5%, at the cost of catching fewer generated clips. Clearing the real-vs-real level does not guarantee replication: of the six feature sets whose GenVidBench score clears it by 0.15 or more, only WaveRep's two keep their score on AIGVDBench. A benchmark score is therefore best read together with that level, a score on a second benchmark, and false-positive rates on other real collections. The toolkit and the clip lists of every evaluation set are at https://github.com/KurbanIntelligenceLab/vidaudit.

cs.CV

版本更新 · v2arXiv:2602.02785

Smell with Genji: Rediscovering Sensory Experience with an AI Co-Smelling Partner

Vera Yu Wu, Awu Chen, Yunge Wen, Yaluo Wang, Jiaxuan Olivia Yin, Yichen Wang, Qian Xiang, Richard Zhang 等 10 位作者

原文摘要

Olfaction contributes substantially to human perception, yet the olfactory-verbal gap often leaves people struggling to articulate what they smell. We present Smell with Genji, an AI-mediated interactive experience adapting Genji-kō, a traditional Japanese scent-matching game. This practice maps comparative judgments across a sequence of scents onto Genji-mon, providing visual patterns as shared reference points for communication. Within this framework, we introduce an AI co-smelling partner integrating multichannel olfactory sensing, a Transformer-based classification model, and a large language model (LLM)-driven conversational interface to translate sensor signals into descriptive dialogue. Rather than deploying machine olfaction purely for identification, the system positions the AI as a co-present partner sharing its sensory impressions while human judgment remains open. Placing the machine's perspective alongside human intuition creates a space for shared reflection, encouraging participants to deliberately evaluate and articulate their choices. This work explores how computational systems can support sensory interaction grounded in dialogue, fostering reflective articulation and preserving human agency in making sense of the olfactory world.

cs.HC

版本更新 · v4arXiv:2609.00038

trajectory-judge:仅看结果的 LLM 评判器在智能体轨迹上遗漏了什么

trajectory-judge: What Outcome-Only LLM Judges Miss on Agent Trajectories

Hadi Mohammadi

AI 速读 基于摘要

作者指出,仅依据请求与最终回复评判智能体轨迹的 LLM 评判器,其召回率可能被高估;应报告相对于干净父轨迹的配对判别力,并按故障是否进入评判器输入、是否改变环境结果分层。

cs.CL · cs.AI · cs.SE

查看原文摘要

A direct test of an LLM judge of agent trajectories injects faults into correct runs and reports recall, per fault type or by whether the fault broke the environment outcome (loud) or not (silent). Such recall can credit a judge with detection it does not have; paired discrimination, its flag rate on the faults minus its rate on the clean runs they came from, exposes this. Our testbed, a deterministic support desk with a scripted oracle and a one-step fault injector, labels all 400 trajectories exactly. A 14B judge shown only the request and final reply scores 34% to 76% recall on four fault types that leave the reply unchanged. There its input is the clean run's, so its paired discrimination is zero and that recall is its flag rate on clean runs. Splitting by outcome survival does not fix this: its loud recall of 84% is a paired +0.393 and its silent recall of 45% a paired +0.048, all from the two fault types that change the reply. Told to check each step, the same model flags every fault of those four types and 0 of 100 clean runs (95% CI up to 3.6%). It does not reliably check the reply: of four invented promises it flags one every time and the other three once in 42 faults. Shown every step but asked only about the reply, it still reaches a paired +0.69 on reply-unchanged faults, against +1.00 when told to check each step. We recommend reporting paired discrimination against clean parents, split by whether the fault reaches the judge's input and by outcome survival, and release the testbed, raw verdicts and analysis pipeline.

版本更新 · v2arXiv:2610.41351

分位数传播规划:面向预算受限的尾部风险评估

Quantile Propagation Planning for Budgeted Tail-Risk Evaluation

Weijia Han, Zhi Huang, Lisha Qu, Yuanyuan Tan, Jiayue Meng, Wenjun Chen, Liying Liang, Bangjie Deng

AI 速读 基于摘要

提出分位数传播规划(QPP),用少量固定扰动分位数经动力学传播并积分代价曲线,在有限模型评估预算下同时评估均值与尾部风险。

cs.LG · eess.SY

查看原文摘要

Risk-aware planning must identify rare, costly outcomes using only a limited number of model evaluations. Monte Carlo rollouts spend this budget on noisy tail estimates, while planning from the mean can miss the tail entirely. We introduce quantile propagation planning (QPP), which propagates a small, fixed set of disturbance quantiles through the dynamics and integrates the resulting cost curve to evaluate both mean and tail risk. The method is exact at the propagated levels, converges as the quantile grid is refined, and admits a one-sided conservatism guarantee under stated structural conditions. In controlled overload and inventory tasks, QPP reduces realized tail cost and constraint violations relative to mean-rollout planning. On overload, it remains competitive with sampled evaluators that use substantially more trajectories. These results position QPP as a transparent, budget-controlled approach to tail-risk evaluation.

版本更新 · v3arXiv:2610.35817

预训练知道做什么,仿真教会怎么做:面向智能体机器人的组合泛化

Pretraining Knows What, Simulation Teaches How: Compositional Generalization for Agentic Robotics

Benjamin Pikus, Daniel Saunders, Mircea Mironenco, Salman Mohammadi, Mathew Mathew, Sirou Zhu, Divya Makkar, Xuhua Huang 等 14 位作者

AI 速读 基于摘要

提出ZAP方法,在仿真中仅用简单任务和模型自身成功经验训练小型开源VLM,使其获得运动控制技能,从而在真实机器人任务上达到与GPT-6 Astra相当的成功率,且更快更便宜。

cs.RO · cs.AI · cs.CV · cs.LG

查看原文摘要

Frontier models such as GPT-6 Astra now directly control robots and can generalize better than vision-language-action models (VLAs). We argue that this ability could arise from two distinct, composable skills: semantic reasoning, to understand what a task asks for in a scene, and motion control, to move the robot to do it. We show that small open VLMs are already capable of the first and mainly lack the second. We introduce Zero-real Agent Policy (ZAP), a recipe that teaches this missing motion skill, entirely in simulation on simple primitive tasks and from the model's own successes, with no frontier teacher or distillation and no real-world robot data. ZAP reaches 96.7% success on real-world tasks, matching GPT-6 Astra in our harness while running 3x faster and 17x cheaper, and 74.0% on a bimanual adaptation of RoboLab-120, on par with GPT-6 Astra's 72.2%. ZAP also generalizes to new objects, goals, and strategies, and in simulation to a robot embodiment on which it never trained.

版本更新 · v3arXiv:2610.17697

LeanPRM:面向自然语言证明验证的 Lean 落地过程监督

LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification

Rodney Lafuente-Mercado, Salman Rahman, Jingyan Shen, James Shiffer, Saadia Gabriel, Pavel Izmailov

AI 速读 基于摘要

LeanPRM 利用 Lean 证明助手自动为自然语言证明生成步骤级标签,训练无需人工标注的步骤验证器。

cs.LG · cs.AI · cs.LO

查看原文摘要

AI systems now write long and difficult mathematical proofs, and checking their step-by-step validity is becoming a bottleneck. A learned verifier could check proofs cheaply, but training one requires a correctness label on every step, and such labels are hard to obtain for open-ended proofs. Existing methods take them from human annotators, which is costly, or derive them from final answers, which proofs do not have, and formalizing each proof in Lean requires expert effort. We introduce LeanPRM, a framework that uses the Lean proof assistant to create step-level labels for natural-language proofs. A model writes a natural-language proof together with a paired formal proof, without feedback from Lean. Lean then labels every step, and a second model discards failures caused only by the formal syntax, so that the negative labels mark errors in the mathematics. LeanPRM contains 10,634 proofs with 152K labeled steps and no human annotation. A Qwen3-4B verifier trained on LeanPRM raises exact step-label accuracy on a held-out set of 200 proofs (LeanPRM-Bench) from 48.2 to 74.7, as opposed to 60.3 when the same base model is trained on PRM800K. It also selects better proofs in best-of-N selection on IMO-level problems, and outperforms the PRM800K-trained verifier on three of four external verification benchmarks. Training on LeanPRM improves verifiers at every size from 0.6B to 8B parameters on LeanPRM-Bench, Hard2Verify and IMO-GradingBench. Given the labels need no human annotation, LeanPRM can grow with the supply of formal statements, offering a path to step-level feedback on AI-generated proofs that are hard for humans to check.

版本更新 · v2arXiv:2610.27439

用可学习图位置编码扩展语言模型中的图推理

Scaling Graph Reasoning in Language Models with Learnable Graph Positional Encodings

Javier Porras-Valenzuela, Arush Arora, Alejandro Ribeiro

AI 速读 基于摘要

作者提出用可学习的、置换等变的图位置编码调制语言模型注意力,并以GNN加Graph Transformer计算编码,从而避免序列化邻接表、把上下文负担降到O(N)。

cs.LG · cs.AI

查看原文摘要

Solving reasoning tasks supported on graphs typically requires providing a Language Reasoning Model (LRM) with a serialized description of the graph. This approach results in the input length growing with the number of edges, which limits the size of the problems that can be solved. Furthermore, even when graphs fit in context, performance sharply decays as the graphs grow. We propose to address this by augmenting LRMs with scalable and permutation equivariant graph positional encodings that modulate attention logits. An additional post-fusion-style module incorporates structural information into hidden states. We compute these positional encodings using a GNN followed by a Graph Transformer. This eliminates the need for serialized adjacencies, reducing the context burden to $O(N)$. We jointly fine-tune the positional encoder and the LRM LoRA adapters using optimal solutions generated by graph-theoretic solvers, resulting in a computationally efficient method. Trained on small graphs, the model exhibits improved size-generalization across graph-theoretic tasks on graphs with up to 1,000 nodes, and produces fewer tokens than a conventional LRM. Additionally, we observe that our model remains robust to node relabelings.

版本更新 · v2arXiv:2610.08701

大幅增长确实会发生:随机矩阵上带部分主元的高斯消元

Large Growth Happens: Gaussian Elimination with Partial Pivoting on Random Matrices

Daniel A. Spielman, Xifan Yu

AI 速读 基于摘要

作者证明对高斯随机矩阵做带部分主元的高斯消元时,出现大幅增长的概率至少是逆拟多项式级别,从而挑战了平滑分析与平均情形分析中的标准猜想。

math.NA · cs.DS · math.PR

查看原文摘要

We prove that the probability that Gaussian elimination with partial pivoting on $n \times n$ Gaussian random matrices has growth $ρ$ is at least inverse quasi-polynomial: $Ω\left(\exp\left(-c \log^2 \left(\fracρ{\sqrt{n}}\right)\log(n)\right)\right)$ for some constant $c > 0$. In particular, the probability of growth $ρ\ge n^{1/2 + ε}$ is at least $\exp(- O (ε^2 \log^3 (n) ))$. This lower bound breaks standard conjectures in the smoothed and average-case analysis of Gaussian elimination. To the best of our knowledge, it is the first non-trivial lower bound on the probability of large growth for Gaussian random matrices.

版本更新 · v2arXiv:2605.07695

OphEdit:无需训练的眼科手术视频文本引导编辑

OphEdit: Training-Free Text-Guided Editing of Ophthalmic Surgical Videos

Ritul Jangir, Arkya Jyoti Bagchi, Mangalton Okram, Saurabh Seetaram Korgaonkar, Aiman Farooq, Deepak Mishra

AI 速读 基于摘要

提出首个无需训练、以文本引导编辑眼科手术视频的框架,通过二阶ODE反演与注意力Value张量注入,在保持眼部解剖结构的同时实现器械替换与手术阶段变化。

cs.CV

查看原文摘要

High-fidelity surgical video generation holds great promise for medical training and AI development, yet precise, controllable editing of such videos remains a formidable challenge. Modifying surgical attributes such as instrument, tissue interactions or procedural phases is particularly difficult due to the strict anatomical and temporal constraints of ophthalmic procedures. We introduce OphEdit, the first training-free framework for text-guided editing of ophthalmic surgical videos. Our approach uses a deterministic second-order ODE inversion pipeline to capture attention Value tensors from the source video and selectively injects them into the conditional branch of classifier-free guidance during denoising, preserving the intricate anatomical geometry of the eye while mapping text-driven semantic modifications onto the video. Clinical evaluation by an ophthalmologist shows that OphEdit handles complex transformations, such as instrument swaps and procedural variations, with superior structural fidelity and temporal consistency compared to adapted general-domain video editors, offering a scalable route to diverse surgical training data. The code and prompts can be accessed at https://github.com/ophedit/OphEdit

版本更新 · v2arXiv:2608.10288

幂律图注意力:缩放点积注意力的精确推广及其在推理阶段的经验性坍缩

Power law graph attention: exact generalization of scaled dot-product attention, empirical collapse at inference

Burc Gokden

AI 速读 基于摘要

PLDR-LLM 的幂律图注意力(PLGA)在理论上精确包含缩放点积注意力(SDPA),但作者报告其头评分算子在推理时趋于输入不变,导致推理坍缩为每头使用常数算子的广义 SDPA。

cs.LG · cs.CL

查看原文摘要

The Large Language Model from Power Law Decoder Representations (PLDR-LLM) and its attention, Power Law Graph Attention (PLGA), replace the fixed bilinear form of scaled dot-product attention (SDPA) with a learned, input-generated bilinear operator $G_{\mathrm{LM}}$, built from a positive tensor $A_{\mathrm{LM}}$ by elementwise power laws. The architecture is fully specified, verified against pinned reference releases; claims are labeled theorem, conditional theorem, measurement, or conjecture. Unconditionally: PLGA contains SDPA exactly at $G_{\mathrm{LM}}=I$; $A_{\mathrm{LM}}$ and $A_{P}$ are strictly entrywise positive, with Perron-Frobenius structure on $A_{\mathrm{LM}}$; the DAG regularizer has the NOTEARS walk-counting form and positivity obstructs exact acyclicity; and, under nonresonance (satisfied by standard rotary frequencies), a commutant criterion identifies which operators preserve relative-position dependence. An inference-collapse theorem: exact input invariance of the head score operators $G_{\mathrm{LM}}$ collapses inference to generalized SDPA with a constant operator per head. Measured invariance: relative fluctuations of $10^{-6}$ and below; perturbation bounds quantify but do not certify cached inference; the assembled proxy misses the decoding margin. A conditional three-stage mechanism (rotary twirl, concentration, row-map contraction) is measured on a released checkpoint. Blockwise training and scoring under the global Gram are stated with explicit target exposure; on tested samples, block and sequential scoring select identical answers and agree on the published TruthfulQA probability-mass metric within $5\times10^{-5}$ per item. Self-organized criticality enters as a phenomenological framework with an intrinsic order parameter; open claims become falsifiable conjectures. Selected proof cores are machine-checked in Lean 4.

版本更新 · v2arXiv:2611.01560

让机器人使用智能体走向大众:一个配开放模型的统一测试框架

Democratizing Robot-Use Agents: A Unified Harness with Open Models

Jiarui Yuan, Yuxin Zuo, Xiaoyu Huang, Wenbo Zhang, Kwanyoung Park, Junlin Yang, Tianxiang Zhang, Ruoqu Chen 等 28 位作者

AI 速读 基于摘要

作者提出统一测试框架 uni-robot-agent,使开放权重模型可开展机器人使用智能体研究并支持后续训练,报告开箱即用的 Qwen3.8-Flash-Next 在 RoboDojo 上取得 31.39 分。

cs.RO

查看原文摘要

Robot-use Agents (RUAs) have shown promising zero-shot manipulation capabilities, but reliance on frontier, proprietary models limits research into further training. To democratize RUA research, we introduce uni-robot-agent, a unified harness for competitive open-weight robot use that supports further training. Without privileged inputs, an out-of-the-box Qwen3.8-Flash-Next achieves a RoboDojo Score of 31.39, surpassing the published GPT-6 Astra-RoboProbe result of 28.97 and many VLA/WAM baselines, and performs competitively on 19 real-world tasks. Our analysis across model generations, families, and sizes links effective robot use to four general capabilities, suggesting that future RUA training should develop them jointly.

版本更新 · v2arXiv:2607.22305

A Roadmap to Impactful Pluralistic Alignment Research

Elinor Poole-Dayan, Jillian Fisher, Jacob Andreas, Mitchell Gordon, Michiel A. Bakker

原文摘要

Much research on responsible AI aims to change the behavior of AI systems that billions of people use. Pluralistic value alignment---the goal of building AI systems that represent and serve diverse human values and perspectives---has emerged as one such active research agenda. Yet, in an audit of the major frontier labs and open-weight developers, we find no public evidence that it has shaped how their models are trained or evaluated. Pluralistic alignment requires adoption to achieve its main goals, so we identify three barriers and lay out a roadmap of research to address each: 1. The justifications for pluralistic alignment so far are mostly normative or theoretical. Studies are needed that measure the benefits and drawbacks of pluralistic AI for users and society. 2. The pluralistic alignment research community has not settled when pluralistic behavior is warranted or what it should look like in practice. Research is needed to specify this behavior concretely enough for developers to operationalize. 3. Existing pluralism evaluations and methods are not ready for deployment. Evaluations and methods are needed that report their effects on other model properties and meet the requirements of production systems. These barriers are not unique to pluralistic alignment, so our roadmap applies to other areas of responsible AI research with similar theories of change. As people hand more decisions to AI systems and AI agents act on their behalf, we hope our roadmap helps this research reach the deployed models making those decisions.

cs.AI

版本更新 · v3arXiv:2609.35711

论多物品拍卖中确定性的力量

On the Power of Determinism in Multi-Item Auctions

Yiannis Giannakopoulos, Johannes Hahn

AI 速读 基于摘要

作者研究单一加性买家、多件独立非同分布物品的垄断拍卖,通过拉格朗日对偶上界改进三种简单确定性拍卖相对最优真实拍卖收益的近似比。

cs.GT · cs.DM

查看原文摘要

We study the classical multi-item monopoly setting with a single additive buyer and $m$ heterogeneous items whose values are independent but not necessarily identically distributed. We compare the revenue (REV) of the optimal truthful auction, which in general may be randomized or complicated, with the revenues of three simple deterministic auctions: selling all items separately (SREV), selling them as a single grand bundle (BREV), and choosing the better of the two. Our technical cornerstone is a general upper bound on the optimal multidimensional revenue, parameterized by freely chosen item-wise caps, and expressed in terms of optimal single-item, Myersonian revenues and basic statistics of the value distributions. We derive this inequality through Lagrangian duality for a nonlinear programming formulation of the worst-case approximation ratio of separate-selling. Combining this inequality with carefully chosen caps and a probabilistic analysis of the resulting terms, we obtain improved approximation guarantees for all three deterministic auctions, both for general independent values and for iid values. Notably, we prove that $REV\le 2.5\max\{SREV,BREV\}$, improving the long-standing $5.2$ factor of Ma and Simchi-Levi and the $6$ factor of Babaioff, Immorlica, Lucier, and Weinberg. We complement this upper bound with a construction establishing a lower bound of $2.19$, improving the bound of $2$ by Rubinstein. For two iid items, we determine the exact approximation ratio of selling separately. Our proof constructs tight Lagrangian dual certificates for the aforementioned programming formulation, when item values lie in the grid $\{0,1/K,2/K,\dots,1\}$, for any discretization parameter $K$. Taking $K\to\infty$ yields the tight ratio $1+W(1/e)\approx1.2785$ in the continuous-valued setting, where $W$ is the Lambert-W function, closing the $[1.278,1.368]$ gap left by Hart and Nisan.

版本更新 · v2arXiv:2601.11420

区间CVaR非线性回归中尾部截断效应的量化

Quantification of Tail-trimming Effects for Interval CVaR Based Nonlinear Regression

Yulei You, Junyi Liu

AI 速读 基于摘要

该文从理论上量化了In-CVaR风险度量对损失右尾的截断在非线性回归抗异常值中的必要性与有效性。

math.OC · cs.LG · stat.ML

查看原文摘要

The tail trimming strategy is widely adopted for robustness against outlier contamination in statistical learning and operations research. However, it generally lacks the thorough understanding that to what extent the tail-trimming technique is effective for nonlinear regression. In this paper, we address the robust nonlinear regression based on the so-called interval conditional value-at-risk (In-CVaR), which is introduced to guard against outliers by trimming extreme losses. We first analyze the trimming effect of the In-CVaR based estimator under distributional perturbation. Our first finding is that the In-CVaR based estimator is qualitatively robust under distributional perturbation in terms of the Prokhorov metric if and only if the largest portion of losses is trimmed. Furthermore, under sample contamination, we derive a probably approximately correct (PAC) bound for the In-CVaR based estimator, showing that the preliminary covariate filtering that is customary in the literature is not needed for In-CVaR based regression. The nonconvexity of the In-CVaR objective, together with unboundedness of covariates, responses and parameter space create the technical challenges for our developed theory, and thus necessitate the new treatment of the qualitative robustness and PAC analysis. Overall, this study rigorously analyzes robustness properties of In-CVaR based nonlinear regression models, which illustrates to what extent the right tail-trimming of the loss by In-CVaR risk measure is essential and effective for robust nonlinear regression. These theoretical findings are further supported by numerical experiments.

版本更新 · v2arXiv:2603.06588

vLLM Hook v1:面向 vLLM 的模型内部状态编程插件

vLLM Hook v1: A Plug-in for Programming Model Internals on vLLM

Ching-Yun Ko, Yunhua Fang, Pin-Yu Chen

AI 速读 基于摘要

作者提出一个开源插件 vLLM Hook,让 vLLM 推理引擎能够按配置捕获或干预模型内部状态,从而支持测试时对齐与增强方法。

cs.LG · cs.CL · cs.PL

查看原文摘要

Modern artificial intelligence models are deployed on inference engines to optimize runtime efficiency and resource allocation, particularly for transformer-based large language models. The vLLM project is a major open-source library to support model serving and inference. However, the current implementation of vLLM limits programmability of the internal states of deployed models. This prevents the use of popular test-time model alignment and enhancement methods. For example, it prevents the detection of adversarial prompts based on attention patterns or the adjustment of model responses based on activation steering. To bridge this critical gap, we present vLLM Hook, an opensource plug-in to enable the programming of internal states for vLLM models. Based on a configuration file specifying which internal states to capture, vLLM Hook provides seamless integration to vLLM and supports two essential features: passive programming and active programming. For passive programming, vLLM Hook probes the selected internal states for subsequent analysis, while keeping the model generation intact. For active programming, vLLM Hook enables efficient intervention of model generation by altering the selected internal states. In version 1 (named vLLM Hook MIA), the plugin runs on vLLM's V2 model runner and preserves the engine's CUDA graphs, so that capturing and steering internal states incurs only a small decoding overhead. It supports both offline use and serving through vllm serve. In addition to presenting the core functions of vLLM Hook, we demonstrate use cases including prompt injection detection, attention-based re-ranking for retrieval-augmented generation, activation steering, hidden-state probing, hallucination detection, schema linking for text-to-SQL, and tool-call risk assessment. Finally, we welcome the community's contribution to improve vLLM Hook via https://github.com/IBM/vLLM-Hook.

版本更新 · v2arXiv:2611.02216

Do agents need to sleep?

Ido Haber, Giulio Tononi

原文摘要

LLM-based agents, which increasingly operate persistently across sessions and accumulate memory through interaction, already exhibit the familiar symptoms of unchecked plasticity: semantic drift, degraded retrieval precision, propagation of stored errors, and saturation of memory capacity. Drawing on the selectionist "ringing, resonance and renormalization" (RRR) account of sleep, we propose that each of its three phases maps onto an operation feasible in digital memory architectures. Engineers building stateful agents, meanwhile, are independently rediscovering the problems sleep evolved to solve: what to forget, and when to do the forgetting. From this we derive the falsifiable prediction that selectionist consolidation will outperform instructionist alternatives as memory stores grow large. The core claim concerns the nature of continuous learning itself: plasticity has a price, and for any system that learns continuously under finite capacity while coupled to its environment, that price is sleep.

q-bio.NC · cs.AI