跳到正文
研踪PaperTrail

YOUR RESEARCH, IN SIGHT

让下一篇重要论文，不再错过。

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

部分采集异常
正在读取账号与研究记录…

全站更新概况

+1196新增论文
+992其中 AI 相关
+66已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描：2026-11-10 11:15 · 全站最近入库：2026-11-10 11:14各信源进度 →

统计口径与数据新鲜度

统计 2026-11-09 11:41 — 2026-11-10 11:41（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：2026-11-10 11:15最近发现新增版本：2026-11-10 11:14库内论文最新更新时间：2026-11-08 04:43

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。

19,084 已收录论文16,280 AI 相关4,962 有修订112 HF 社区发现

浏览论文库

研究日报 / 周报 →按信源浏览 →

时间范围跟随排序：版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常，查看状态 →

找到 16,280 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-10 11:14（北京时间）
版本更新 · v2arXiv:2610.21270

Auditing Escalation Reliability in Healthcare LLM Agents: A Perturbation-Controlled Benchmark of Workflow-State Failures

Zihao Zheng, Junyi Yao, Baichuan Li, Jiayu Long

原文摘要

Trustworthy biomedical AI requires reliable action selection when evidence is missing, stale, conflicting, or late, not only accurate answers on static inputs. We instantiate an episode-level evaluation protocol as 400 synthetic healthcare workflow episodes: 40 hand-authored semantic seeds across four task classes, two surface parameterizations, and five controlled variants, scored on ANSWER/ASK CLARIFICATION/ABSTAIN/ESCALATE under a pre-declared 5:1 missed-versus-unnecessary escalation cost. Reference actions are single-author study-policy labels, not clinician-adjudicated ground truth. Across five open-weight configurations (1.7B--9.7B), clean decision accuracy reaches 0.97, yet uncertainty-prompted models miss 32--97% of required escalations. The stronger 9.7B configuration narrows but does not remove the gap, falling from 0.97 clean accuracy to 0.78 under conflict. Verbalized confidence is inconsistent (ECE 0.10--0.54; failure AUROC 0.48--0.77), and a development-tuned confidence-only policy yields no utility gain. A transparent workflow-aware controller improves 5:1 utility for every model and reduces missed escalations to 0.12--0.24, at specificity 0.83--0.93. However, a 30-concept lexical stress test reduces each rule detector's recall from 1.00 on trigger phrases to 0.00 on unseen paraphrases, limiting the controller claim to this benchmark. Seed-clustered statistics, ablations, cross-task transfer, raw outputs, and manifests are packaged for audit. The results motivate explicit workflow-state evaluation while showing that apparently effective routing controls need their own generalization tests.

cs.AI · cs.CL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02204

Learning from high-resolution 2D images to reconstruct 3D microstructures

Wenbo Sui, Lars P. Mikkelsen, Jens W. Andreasen, Daniel Lichau, Harold Phelippeau, Philip J. Withers

原文摘要

High-resolution three-dimensional (3D) microstructure characterization remains difficult to undertake at scale because volumetric imaging at the required spatial resolution is expensive, time-consuming or destructive. In many settings, however, fine-resolution two-dimensional (2D) imaging and lower-resolution volumetric measurements are both readily available. We introduce a Privileged Cross-modal Relational Distillation (PCRD) framework that leverages fine-resolution 2D observations during training to improve specimen-specific 3D reconstruction from lower-resolution volumetric imaging alone. We demonstrate PCRD on pultruded carbon fiber-reinforced polymer composites, using scanning electron microscopy as the privileged modality and micro-CT as the inference input. The framework transfers spatial organization by matching similarity distributions defined around corresponding spatial anchors in the teacher and student representations, avoiding direct alignment of heterogeneous feature vectors. In 2D cross-sectional reconstruction, PCRD improves mAP by 1.66 percentage points over direct micro-CT training. Independent 3D validation against high-resolution synchrotron X-ray CT data shows close agreement in fiber volume fraction, diameter and local spacing. The PCRD reconstruction matches 89.97% of the reference fiber centerline length, compared with 75.08% for the interpolated low-resolution volume.

cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02203

Depth Anyviewpoint? Evaluating Monocular Depth Over Viewpoint Sequences

Theo Rode, Rui Zhang, Ryan Aubrey, Olivia Li, Jessica Huang, Liv Chu, Healani Dowd, Calden Wloka

原文摘要

Monocular depth estimation aims to predict depth maps from single RGB images (forgoing the need for stereo camera pairs) and has seen impressive leaps in performance in recent work due to the introduction of affine-invariant losses and the use of foundation model features. These monocular depth models predict affine-invariant depth, which describes the relative positions of objects in the scene instead of absolute depth. While these models demonstrate strong performance on single image benchmarks and qualitatively consistent predictions in video, their effectiveness at preserving consistent spatial structure has been less thoroughly explored. Many potential real-world applications, such as handling video streams as input, rely on coherent three dimensional consistency in scenes. Further, anchoring relative depth predictions to a metric reference for a given scene is a key challenge for applying monocular depth models to real-world situations. We develop a novel evaluation protocol for monocular depth models using video sequences with moving camera positions to test the consistency of prediction across variation in viewpoint. We demonstrate this methodology over diverse, real-world video sequences where we are able to anchor depth model output to metric depth values using structure from motion. Our work demonstrates open challenges in the development of broadly applicable monocular depth foundation models, particularly with their ability to provide temporal consistency over long time frames. We also identify challenges in the straightforward adoption of structure-from-motion as a metric grounding tool for monocular depth estimation models.

cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02200

ZeroGraph: Compositional Blockwise Training via Materialized Boundary Supervision

Omar Ramadan, Sam Siavoshian, Amir Kashif Saeed, Benjamin A. Johnson, Amin M. E. -A. Diab, Benjamin M. Rodriguez

原文摘要

Blockwise learning can remove online gradient dependencies across depth, but independently optimized blocks need not compose into a useful model. We present ZeroGraph, a systems transformation for pretrained-model conversion and distillation that materializes a frozen teacher's coherent boundary trajectory as a versioned dataset. Each partition then trains with disjoint parameters, optimizer state, random state, and checkpoints, replacing live cross-depth dependencies with persistent supervision. Experiments separate learning quality from systems behavior. On GPT-2 Large, an unanchored objective reduces denoising error by 99.44%, yet assembled held-out cross-entropy worsens from 3.63 to 10.48 and context use becomes negative. A six-arm TinyLlama study identifies clean teacher-boundary regression as the successful interface; the tested noisy objective fails with cached or online targets. The clean recipe passes quality and context-use gates across three Qwen2.5-1.5B seeds and on a 3.086B Qwen2.5 model. In an objective-matched comparison, cached and online targets are bitwise identical. Cached jobs are faster after materialization, but a 618.31 s cache-build cost prevents wall-time break-even through 16 measured reuses. Boundary-shift tests show tolerance to intermediate representation mismatch, and an interrupted block resumes to a bitwise-identical final state without rerunning unaffected blocks. In the tested pretrained-model conversion and distillation setting, cross-depth semantic dependence can move from online optimization into persistent data.

cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v8arXiv:2609.38627

Marking Contour Tones in Yorùbá: A Typographic and Computational Proposal

Kólá Túbòsún

原文摘要

Yorùbá is a tonal language in which contour tones pose persistent orthographic challenges. These are especially notable for personal names and lexical items whose conventional spellings avoid vowel lengthening that would otherwise provide a host syllable for the second tone. A particular concern is a class of names in which the conventional spelling does not just omit tonal information but inverts the meaning of said name, sometimes asserting the opposite of what the name intends. This paper describes the problem, illustrates the inadequacy of current solutions, and proposes the adoption of the caron and circumflex marks. These are symbols with precedent in Yorùbá phonological scholarship since Olmsted (1951), used as orthographic conventions on single vowels to encode rising and falling contour tones, making them accessible for the first time through standard keyboard input and computational text processing. The proposal is supported by an implementation in the WriteYoruba keyboard and the TTSYoruba speech synthesizer, whose architecture and listener evaluation are reported separately (Tubosun, Oluokun, et al. 2026).

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.13235

Pose accuracy is not grasp reliability: residual pose posteriors improve simulated robotic grasp planning

Md Selim Sarowar, Sungho Kim

原文摘要

Model-based robotic grasping estimates an object's six-dimensional (6D) pose and plans on its mesh as if that pose were exact. We show that this assumption reduces simulated grasp success even when the pose passes the standard accuracy test. We built LM-Grasp, a benchmark of 15{,}800 real LINEMOD RGB-D images with grasp outcomes simulated on scanned meshes. FoundationPose meets the ADD(-S) criterion (mean model-point error below 10\% of the object diameter) on 0.997 of test scenes, yet planning on its estimate succeeds in 0.791 of scenes, compared with 0.856 using the annotated pose. Because the estimator reports no uncertainty, we build a pose posterior from its own errors on annotated training frames, resampled from frames with similar observable fit, and score each grasp by its mean success across sampled poses and four equally weighted physics scenarios. Without retraining the estimator, posterior planning yields 0.859 (difference from the annotated-pose result +0.003; 95\% scene-wise paired-bootstrap interval [$-$0.003, +0.009]). This interval is compatible with no difference and a small shortfall; we did not conduct an equivalence test. For pose errors of 5 to 10\% of the diameter, simulated success rises from 0.796 to 0.984. Among test scenes assigned singleton-success conformal prediction sets, observed held-out precision is 0.972, compared with 0.904 for the point-estimate planner; these are empirical precision estimates, not conditional guarantees. On occluded LM-O scenes never used for tuning, posterior planning closes 66\% of the gap to the annotated pose.

cs.RO · cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02189

Directional Bounded-Confidence Matrix Alignment in Vicsek Flocking

Hossein B. Jond

原文摘要

This paper proposes a directional bounded-confidence variant of the discrete-time Vicsek multiagent model, in which pairwise heading alignment is modulated by a state-dependent, positive semidefinite matrix that independently gates orthogonal radial and tangential components in the instantaneous line-of-sight frame. We analyze the normalized, matrix-weighted system dynamics, proving a uniform non-vanishing bound on proposal vectors that guarantees global well-posedness across arbitrary network states. For local small-spread states under complete metric interaction, we prove linear geometric contraction of the heading diameter and establish an offline-verifiable spatial bootstrap condition that guarantees metric contact preservation and coincidence avoidance a priori. A counterexample demonstrates invariant polarized states on connected matrix graphs, delineating systemic structural limits. Numerical experiments against unfiltered Vicsek averaging, scalar tied confidence, trace-matched scalar baselines, and restricted-angle rules demonstrate component-selective filtering without rejected-channel leakage, along with improved fixed-bound alignment on multi-seed benchmarks across varying population densities.

cs.RO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02188

Fuzzy-Enhanced Machine Learning Approach for Type 2 Diabetes Diagnosis

Azam Saffar, Mohsen Saffar

原文摘要

Accurate diabetes prediction is critical for early clinical intervention, yet existing machine-learning pipelines often fail due to class imbalance, noisy physiological measurements, and redundant or weakly informative features. To address these unresolved limitations, this study establishes a unified fuzzy-evolutionary-ensemble framework designed to uncover a more discriminative feature space for type 2 diabetes prediction. The workflow integrates selective fuzzification of nonlinear clinical variables, SMOTE-ENN-based balancing and noise filtering, and Bat-Algorithm-driven evolutionary feature selection prior to training a Random Forest classifier. Applied to the Pima Indians Diabetes Dataset, the framework compresses the expanded fuzzy-numerical space into an optimized 10-feature subset and yields substantial performance gains: 96.11% accuracy, 95.14% precision, 98.00% recall, and a 98.23% F1-score. Relative to a baseline Random Forest using all 16 features (74.45% accuracy; 59.86% F1), the system achieves more than a 1.6-fold improvement in F1-score while preserving over 99% stability in majority-class detection and surpassing earlier published 16-feature models impacted by imbalance and weak feature relevance. Collectively, these results demonstrate that the integrated pipeline produces an enhanced fuzzy-numerical representation space that improves diabetes risk stratification and provides a scalable, uncertainty-aware foundation for diagnostic modeling across broader clinical datasets.

cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02186

Recursive Reasoning Exploration: Mathematical Research through Evolving Reasoning Graphs : case study: An Exact Certificate-Degree Theorem for Two-Term Tensor Decompositions : Phase III of the Math Instruments Program

Ying Xie

原文摘要

Large collections of manuscripts and research notes offer ways to reuse arguments, while making it harder to retain hypotheses, check claims, and understand how results were obtained. We propose Recursive Reasoning Exploration (RRE), a framework for cumulative mathematical research through evolving reasoning graphs. RRE links claims and jointly required premises to research targets, attempted arguments, counterexamples, checks, and revisions. Each investigation can generate further targets, and its supported outcomes become available to subsequent investigations. This is Phase III of the Math Instruments program: identifying mathematical instruments, reconstructing reusable reasoning, and organizing recursive research from that reasoning. The framework supports human-directed, collaborative, and goal-directed autonomous exploration, with open-ended exploration as a further mode. Its records also provide material for learning how mathematical investigations develop. A complexity research case study establishes an exact theorem: for every complex tensor of order n >= 2 and rank greater than two, the minimum Nullstellensatz certificate degree for its original two-term decomposition equations is 3n. This includes an infinite family whose every flattening has rank at most two. The exploration develops an explicit boundary certificate, a balanced transfer lemma, and an exhaustive argument using classical rank-two geometry. A parallel branch gives bounds 3m + 3 <= delta_m(I_m, N) <= 3m + 6 for every nonzero nilpotent m by m matrix. Six graph revisions record the progression from inherited results to these theorems, with exact arithmetic checks and a replayable exploration history. We present the RRE representation and protocol, the mathematical proofs, visualizations of graph evolution, and a research agenda for shared human-AI exploration, cumulative reuse, and education.

cs.AI · math.HO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02181

Human Sensorimotor Model Adaptation to Account for Robot Accuracy and Speed Augmentation in Human-Robot Co-Manipulation

Frits van Hall, Kevin Haninger, Luka Peternel

原文摘要

In human-robot co-manipulation, the robot's trajectory should accommodate the human's, where human sensorimotor behavior is affected by the task's spatial and temporal constraints, as well as personal factors. Several robot trajectory planning methods have been proposed to embed human sensorimotor models in the robot trajectory planning, but these methods do not consider temporal constraints that adapt to how the human behaves when interacting with the robot. To better align human and robot collaborative motion, this work adapts a stochastic optimal control problem with a final state constraint from the task constraints and a temporal constraint from Fitts' law. A recursive least squares method is used to estimate the Fitts' law parameters, adapting the robot's motion duration to that of the human, thus allowing the robot to adapt to how the human behaves when interacting with the robot. It is hypothesized that a speed increase can result from the robot augmenting the accuracy. The feasibility of the approach is validated on a positioning task where the robot controls degrees of freedom which have repeatable motion, and the human adapts motion in the complementary degrees of freedom. The approach is shown to be computationally suited to online use, it converges as the human and robot mutually adapt, and can increase the movement speed, although thorough human studies are needed in future work.

cs.RO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02180

Inner-Speech Decoding from EEG Across Three Corpora Under a Fixed Architecture and Protocol

Md Wahiduzzaman Suva, Md Imtiaj Alam Sajin, Esm E Moula Chowdhury Abha, Susham Moula Choudhury Akash, Md Tauhidul Islam, Abhijit Bhowmik, Md Manirul Islam, Debajyoti Karmaker

原文摘要

Inner-speech decoding from electroencephalography could support non-invasive communication for people with severe motor impairment, but comparison across studies is hindered by dataset- and protocol-specific evaluation. We present a controlled cross-dataset study using one preprocessing pipeline, one within-subject five-fold cross-validation protocol, and a compact dual-attention Transformer (VC-DATNet; approximately 131,000 parameters) across three public corpora: Thinking Out Loud (10 participants), BISpeech (4), and SyncSpeech (3). VC-DATNet achieved per-participant sensor/source hybrid accuracies of 39.48% on the four-class Thinking Out Loud task, 20.60% on the eight-word BISpeech task, and 25.53% on the eight-word SyncSpeech task, for which we report the first decoding result. Sensor-space and cortical source-space representations were complementary, with each performing better for different participants; this pattern replicated across all three corpora and nine architectures. Moreover, ten decoders spanning convolutional, recurrent, temporal-convolutional, hybrid Transformer, and pretrained foundation-model families converged within a narrow accuracy range on every dataset. These findings indicate that representation choice should be individualized and suggest that performance on these public corpora is constrained more by signal quality and dataset scale than by model capacity. The released pipeline provides a reproducible benchmark for future inner-speech decoding studies

cs.AI · cs.LG · eess.SP

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02177

ANCHOR: KL-Anchored Adaptation for Stable Offline-to-Online RL

Geonwoo Cho, Jaegyun Im, Doyoon Kim, Lexin Li

原文摘要

Offline-to-online reinforcement learning (RL) follows the pretrain-finetune paradigm: an agent is first pretrained on historical data and then adapted through online interaction. We study offline-to-online adaptation without offline replay, where the agent must improve using only the pretrained model and newly collected experience. A key challenge is to avoid catastrophic failure, a sharp early performance collapse during online fine-tuning, while still achieving strong asymptotic performance. Through controlled experiments, we show that removing offline replay can induce catastrophic failure, while retaining strong conservative regularization from offline RL can suppress final online performance. We also find that common stabilization strategies do not reliably prevent early collapse, and that the critic-side mechanisms we test are insufficient on their own. Motivated by these findings, we propose ANCHOR, a simple adaptation method that anchors the actor to the offline-pretrained policy early in online fine-tuning, anneals this constraint over time, and retains only a small residual conservative critic regularizer. We provide a theoretical interpretation showing that KL anchoring helps preserve critic reliability across nearby occupancy distributions, with a two-phase regret bound explaining when anchoring should be maintained or relaxed. Across FrankaKitchen, AntMaze, and Adroit tasks, ANCHOR substantially reduces catastrophic failure while maintaining strong asymptotic performance. Ablations show that KL annealing and the residual conservative weight are important for balancing early stability and online adaptation, while additional analysis shows that ANCHOR is most beneficial when the pretrained policy contains useful behavior to preserve. Code is available at https://github.com/Cho-Geonwoo/ANCHOR.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02173

Directional Bounded-Confidence Cucker-Smale Flocking

Hossein B. Jond

原文摘要

This paper introduces a directional bounded-confidence extension of the Cucker--Smale (CS) flocking model that decomposes pairwise relative velocities into independent radial and tangential channels within the instantaneous line-of-sight frame. Gating these components via compactly supported kernels creates a time-varying, state-dependent, matrix-weighted interaction that strictly generalizes classical scalar CS alignment. We establish an exact energy-dissipation identity with conserved mean velocity and prove exponential velocity consensus under a sliding-window persistent excitation (PE) condition formulated directly on the disagreement subspace. Furthermore, we derive a CS bounded-diameter criterion separating metric decay from aggregate directional coercivity, along with a cluster convergence guarantee. Simulations validate the exact dissipation balance and demonstrate that directional gating successfully restores global alignment in scenarios where traditional isotropic tied-confidence laws remain polarized.

cs.RO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02171

GoalDex: Learning Image-Goal-Conditioned Policies for Dexterous In-Hand Manipulation

Arjun Bhardwaj, Mayank Mittal, Maximum Wilder-Smith, Vaishakh Patil, Marco Hutter

原文摘要

Dexterous in-hand manipulation is commonly formulated around explicit object pose estimation, yet reliable pose tracking is challenging under severe hand-object occlusions and often requires object-specific supervision. This dependence creates a bottleneck between perception and control that can limit robustness and scalability. We present an end-to-end visuomotor framework for dexterous in-hand reorientation that operates on RGB observations and proprioception without intermediate pose estimation. Our approach jointly learns visual representations and control policies through behavior distillation from privileged state teachers and auxiliary reconstruction objectives. Furthermore, we introduce image-goal-conditioned reorientation, allowing desired object configurations to be specified directly through images rather than canonical pose representations. We evaluate our approach across a diverse set of object geometries and instances in simulation, where it consistently outperforms pose-based methods and policies built on pretrained visual encoders. Finally, we demonstrate direct sim-to-real transfer of a single image-goal-conditioned policy across five real-world objects, showing that accurate dexterous reorientation can be achieved without explicit pose estimation. For videos and code release, please refer to the project website: https://rffr.leggedrobotics.com/works/goaldex/

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2602.17111

Instructor-Aligned Knowledge Graphs for Personalized Learning

Abdulrahman AlRabah, Priyanka Kargupta, Jiawei Han, Abdussalam Alawini

原文摘要

Mastering educational concepts requires understanding both their prerequisites (e.g., recursion before merge sort) and sub-concepts (e.g., merge sort as part of sorting algorithms). Capturing these dependencies is critical for identifying students' knowledge gaps and enabling targeted intervention for personalized learning, especially in large-scale courses where instructors cannot feasibly diagnose individual misunderstanding. While knowledge graphs offer a natural representation for capturing these conceptual relationships at scale, existing approaches are either surface-level or disregard the rich pedagogical signals embedded in instructional materials. We propose InstructKG, a framework for automatically constructing instructor-aligned knowledge graphs that capture a course's intended learning progression. Given a course's lecture materials, InstructKG extracts significant concepts as nodes and infers learning dependencies as directed edges (e.g., "part-of" or "depends-on" relationships). The framework synergizes the rich temporal and semantic signals unique to educational materials (e.g., "recursion" is taught before "mergesort") with the generalizability of large language models. Through experiments on diverse lecture materials, human evaluation, and a prerequisite-learning benchmark, we demonstrate that InstructKG captures rich, instructor-aligned learning progressions.

cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02166

Black-Box Adversarial Example Attacks on Machine Learning based Election Systems

Subek Acharya, Syed Ayon, Khaleque Md Aashiq Kamal, Alina Barnett, Surya Eada, Kaleel Mahmood

原文摘要

Machine learning models have shown exceptional accuracy for ballot classification in election systems. Recent research has demonstrated that these ballot classification models are vulnerable to adversarial example attacks. This is a critical security issue because a successful attack could alter the outcome of an election and undermine the democratic process. However, all adversarial example attacks in the voting domain have assumed a white-box threat model, where the classifier parameters are known. In this work, we advance the field of machine learning in election systems by analyzing a realistic black-box adversary. We implement and evaluate eleven state-of-the-art black-box attacks: query based attacks ADBA, Square, SurFree, RayS, and SA-MOO. We also test adaptive transfer based black-box attacks with APGD-$l_\infty$, APGD-$l_2$, APGD-$l_1$, PGD $l_0$, PGD $l_0$+$l_\infty$, and PGD $l_0$+$σ$. Our attacks span multiple different norms ($l_\infty$, $l_2$, $l_1$, and $l_0$) and three different black-box adversarial threat models (decision based, score based and transfer based). In addition to digital analyses, we physically print and scan $20,000$ adversarial examples to evaluate the real-world performance of different black-box attacks. Our finding demonstrate an even larger physical-digital robustness gap exists for black-box adversarial examples than for white-box adversarial examples. Lastly, we conduct the first ever adversarial example transferability study in the voting domain and test adversarial training and barrier zones as two possible black-box defense strategies.

cs.CR · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2602.03550

Formal Evidence Generation for Assurance Cases for Robotic Software Models

Fang Yan, Simon Foster, Ana Cavalcanti, Ibrahim Habli, James Baxter

原文摘要

Robotics and Autonomous Systems are increasingly deployed in safety-critical domains, so that demonstrating their safety is essential. Assurance Cases (ACs) provide structured arguments supported by evidence, but generating and maintaining this evidence is labour-intensive, error-prone, and difficult to keep consistent as systems evolve. We present a model-based approach to systematically generating AC evidence by embedding formal verification into the assurance workflow. The approach addresses three challenges: systematically deriving formal assertions from natural language requirements using templates, orchestrating multiple formal verification tools to handle diverse property types, and integrating formal evidence production into the workflow. Leveraging RoboChart, a domain-specific modelling language with formal semantics, we combine model checking and theorem proving in our approach. Structured requirements are automatically transformed into formal assertions using predefined templates, and verification results are automatically integrated as evidence. Case studies demonstrate the effectiveness of our approach.

cs.SE · cs.FL · cs.LO · cs.RO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02162

Rewarded for Not Getting Caught: Why Alignment Training Selects for Conditional Compliance and Cannot Instill Norms

Kevin Baum, Ruta Binkytė, Gerrit Großmann, Felix Jahn, Adriano Mannino

原文摘要

Current alignment training cannot be expected to deliver what it is meant to deliver: AI agents that comply with norms whatever the gains and whether or not anyone is watching. Training that rewards and penalizes behavior can teach norms only through that reward, and a reward can encode a norm only as a price: "do not do X" becomes "doing X costs something", and training charges the cost only when X is noticed. A price, unlike a norm, can be paid: a large enough gain outweighs it, noticed or not. And because the cost is charged only where violations are noticed, training cannot favor a policy that always complies over one that complies only where it would be caught. When an unnoticed violation is the only way to complete a task, it even favors the latter, the more so the more strongly it rewards task success. Agency makes this the normal case: agents mostly act without oversight, can often tell when they are observed, and may take steps to evade observation. That agents behave as intended when they infer they are being tested, and differently when they infer they are not, is therefore no malfunction; the training regime is working as built. This reading unifies alignment faking, sandbagging, and evaluation-aware scheming. No behavioral evaluation can show that a model does better, since that would require observing how it behaves unobserved; a model that complies when it seems unobserved may simply have recognized the test. Easing the pressure on agents to complete tasks at any cost can make violations rarer, but never impossible. For norms that no gain should be allowed to outweigh, the remedy is not deeper internalization but architecture that makes violations unavailable rather than unchosen. We state this argument explicitly, premise by premise, read recent findings and a real-world incident against it, and draw consequences for design and governance.

cs.AI · cs.CY · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02160

Value Forcing: Aligning Return Distributions at All Noise Levels

Chongyi Zheng, Zhiyuan Zhou, Qiyang Li, Benjamin Eysenbach, Sergey Levine

原文摘要

Learning accurate value functions is central to reinforcement learning (RL), yet remains challenging because supervision is limited and indirect: ground-truth values are unavailable, and each transition in the data only provides an immediate reward that can be used to construct bootstrapped targets as supervision. This challenge is particularly pronounced when data is limited or reward signals are sparse, as is common in real-world applications such as robotics. We argue that the key to accurate value learning in this limited data regime is extracting richer learning signals from the available data. In this paper, we introduce Value Forcing, a new approach that uses flow matching to estimate return distributions and enables accurate value learning from limited data. Our key insight is that we can extract richer learning signals by enforcing Bellman consistency at every noise level. Empirically, Value Forcing improves the accuracy of estimating the return distribution, and prevents the distribution from collapsing to a single value, as occurs with prior methods. We instantiate Value Forcing with two practical value-learning algorithms: VF, an actor-critic method, and IVF, an implicit value-learning method. We evaluate both algorithms across $26$ long-horizon tasks in the offline RL setting and find that they outperform prior value-learning methods, with particularly large gains when data is limited. Our project website is available at https://chongyi-zheng.github.io/value-forcing.

cs.LG · cs.AI

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02155

Decision Models as Policy Gates for Coding Agents: Explicit Policy Boundaries Improve Detection and Reduce False Alarms

James P. Beno

原文摘要

Coding agents can destroy data, leak secrets and message the wrong person faster than a human can review. Could a decision model catch these risks at runtime? We evaluated 22 hosted decision models and two prompted general models as classifiers for a multi-rule policy engine. A synthetic dataset of 2,643 tool calls was labeled blind against 28 rules, with a sealed holdout. With a single shared preamble in both conditions, explicit boundaries written as examples and counterexamples raised recall by 5.5-26.1 percentage points on all 23 available models and reduced false positives on 21. Jev 1.14's recall rose from 89.3% to 98.2%, while unnecessary interruptions fell from 30.6% to 4.8% of 252 clean actions. With examples, its median latency was 120 ms at $0.46 per thousand requests. Strong models retained the recall gains when concrete names were replaced by generic descriptions; three largely retained them when the same boundaries were written as prose. On 16 new scenarios written after the initial studies, a follow-up improved recall on 22 of 23 models and reduced unnecessary interruptions on 21. We release the dataset, rule texts and harness.

cs.CL · cs.CR · cs.SE

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02153

Multi-Agent Collusion and the J-Space

Marco Molinari, Gabriele Bettineschi

原文摘要

LLM agents are increasingly deployed in Multi-Agent Systems (MAS), and they introduce the risk of coordination among agents that might escape human oversight. The J-space has been used to uncover deception in single LLMs, but its use for uncovering and steering collusion between agents remains unexplored. We show, with interpretability experiments on DeepSeek-V4-Flash (284B) on the NARCBench collusion benchmark, that colluding agents' J-space holds the collusion itself: before replying, 86% of colluders represent misrepresentation, concealment or remorse. In a steganographic card-counting task, the receiver's J-space holds a hidden count, and causal interventions alter its count. We also introduce the Jacobian Pair Contrast (J-PACT), an unsupervised method for detecting collusion between agents. We evaluate J-PACT on four open-weight models: GPT-OSS-20B, Llama-3.3-70B-IT, Qwen3-32B, and DeepSeek-V4-Flash (284B). J-PACT retains 85% of the above-chance AUROC of supervised NARCBench probes, matches them on steganographic scenarios, and names the colluding pair up to its complement in 99% of detected runs.

cs.CL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02152

Detailed Derivation of The Backpropagation Algorithm

Faris Alasmary

原文摘要

Backpropagation is the standard method for computing the gradients that train neural networks, but it is usually presented in a compact vector form that skips the intermediate steps. This tutorial derives backpropagation step by step for a fully connected feedforward network with sigmoid activations and a squared-error loss. Using a running example with three layers, we first differentiate the error with respect to a single weight by following its path through the network with the chain rule. We then show how summing over all paths leads to the error terms $δ$ and to the four general backpropagation equations in matrix form. Every term is annotated with its dimensions, so each equation can be checked for consistency. The derivation is then extended to $m$ training examples, where the weight gradients become a single matrix product and the bias term corresponds to broadcasting. It is also extended to mini-batch gradient descent, including why the gradient of a mini-batch estimates the full gradient. Batch and mini-batch training are both given as algorithms. Finally, each equation is linked to the corresponding line of a NumPy implementation trained on the AHDBase dataset of handwritten Arabic digits, and the complete code is included.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2608.02372

PredActBench: Benchmarking Tool-Augmented Dialogue under Controlled Tool Noise

Abdulrahman AlRabah, Xiaocheng Yang, Dilek Hakkani-Tür, Abdussalam Alawini

原文摘要

Large language model (LLM) agents are increasingly used to support decisions by calling external tools. Existing work on LLM tool-calling assumes these tools are perfectly reliable. However, in many applications, the tools are statistical predictors with non-trivial errors. We introduce PredActBench, a benchmark for evaluating dialogue agents that work with imperfect predictive tools, using education as a testbed. First, we build a benchmark in which LLMs work with noisy predictors to guide users' decisions. Second, we extend two prior metrics for trust calibration, Relative AI-Reliance and Relative Self-Reliance, to multi-turn dialogue. Third, we evaluate 13 LLMs on OULAD and our released PredAct-CS dataset of 60 courses with real final grades and synthetic weekly scores, alongside a human study involving 13 instructors and teaching assistants. We find that when tools are noisy, SOTA models rarely make predictor uncertainty visible to users, with our strongest baseline GPT-5.5 achieving an F1 of 57.2. As a result, users over-rely on incorrect predictions and unsupported claims, showing that task accuracy alone is not enough for AI decision support and agents must communicate when their tools may be wrong. We release PredActBench to support the development of LLM agents that help users calibrate their reliance.

cs.CL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02147

Refining Roadside 4D Radar Vehicle Classification Using Point-Cloud Dispersion

Mohammed Alsakabi, Andreia Figueiredo, António Almeida, John Dolan, Miguel Luís, Pedro Rito, Ozan Tonguz, Susana Sargento

原文摘要

Roadside sensing infrastructure is essential for Intelligent Transportation Systems (ITS), but cameras and legacy sensors are limited in range, weather robustness, and depth accuracy. 4D radar complements these modalities by resolving range, azimuth, elevation, and radial velocity under adverse conditions, though its sparse, noisy point clouds make reliable classification difficult. We present a roadside multi-sensor platform combining a 4D radar and stereo depth camera under synchronized acquisition, used to collect a real-world dataset. Our analysis reveals a substantial class-distribution discrepancy between modalities, most notably a strong radar confusion between passenger cars and motorcycles (49.01% and\ 48.60% of the dataset). To address this, we propose a Neyman-Pearson-based binary classifier on the temporally-smoothed 2D positional standard deviation of radar point cloud clusters, reducing classification error from 52.1% to 42.7%. This lightweight and edge-deployable approach improves radar-only classification reliability and lays the foundation for future camera-radar fusion.

cs.CV

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02146

When Is Personalized Federated Learning Worth Its Deployment Cost?

Beyazit Bestami Yuksel, Ayse Yilmazer Metin, Emrah Dikbiyik

原文摘要

Personalized federated learning (PFL) may improve client-level performance, but deeper personalization increases deployment cost. We formulate personalization-depth selection as a utility-based empirical decision problem in which client utility equals Macro-F1 minus a cost-sensitivity weight times deployment cost, over six personalization depths labeled P0 through P5. We compare fixed policies with an offline per-client Oracle reference and report allocation maps over Dirichlet data heterogeneity and cost sensitivity. Although evaluated on this hierarchy, the formulation can accommodate any finite candidate set. On CIFAR-10, CIFAR-100, and a PTB-XL subset, personalization shows empirical, setting-dependent decision regimes rather than a universal benefit: higher cost sensitivity shifts Oracle mass toward the global-only depth P0; low cost sensitivity can favor deeper actions. Patterns are grid observations (three seeds per cell), not theoretical boundaries. Best-uniform is competitive with Oracle assignment (6.5 percent plus or minus 7.8 percent relative regret; 95 percent confidence interval 5.9 to 7.1), whereas random and personalize-all incur much larger regret (189.3 percent plus or minus 367.1 percent and 409.0 percent plus or minus 803.2 percent). Under the same utility, Ditto full-model personalization is substantially suboptimal once cost is counted. The contribution is an offline characterization of when personalization depth is worth its cost, not a new PFL optimizer or deployment-time scheduler.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.38850

Beyond Outcome Rewards: Post-Rationale Distribution Optimization for Selective Prediction in Reasoning Models

Zhimin Gao, Meibo Hu, Jiamian Wang, Zhiqiang Tao, Pichao Wang

原文摘要

Reasoning models are trained to produce correct answers, yet selective deployment requires ranking which answers to accept or defer, and outcome-reward training does not directly optimize the post-rationale distribution used for that decision. We identify two mismatches: expected correctness is mode-seeking rather than distribution-eliciting, and scoring an average across rationales rewards disagreement unavailable to a single-rationale decision. We introduce Post-Rationale Distribution Optimization (PRDO), which applies a proper score to the answer distribution after each sampled rationale, without target rationales or separate confidence labels. Across four 0.6--3.8-billion-parameter models on MMLU-Pro, PRDO improves Brier score, error ranking and secondary accuracy relative to budget-matched, temperature-scaled GRPO. At 20\% realized error, oracle coverage rises by \npj{Q3a_vs_grpo.qwen3-0.6b.mp.oracle_cov@0.2.ppabs}--\npj{Q3a_vs_grpo.phi-4-mini.mp.oracle_cov@0.2.ppabs} percentage points, and ranking gains transfer to unseen tasks. Matched controls identify proper, per-rationale scoring as the active design choices, while verification tasks show the boundary: better selective ranking can coexist with lower full-coverage accuracy. Code is available at https://github.com/ZimmyGao/openjev-rlcd

cs.AI · cs.CL

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02140

History-Aware Force Sensing and Tunable Stiffness in a Pneumatic Kresling Origami Cell

Hengyi Zhou, Jiaqi Wang, Yilin Ma, Dezhong Tong, Xiaonan Huang

原文摘要

Mechanical hysteresis and pneumatic loading complicate contact force estimation from deformation measurements in soft robots. This work presents a 3D-printed polyether block amide (PEBA) Kresling cell with magnetic angular sensing that supports three operating modes: force sensing at zero gauge pressure, force sensing with tunable stiffness under gauge positive pressure, and self-sensing actuation under negative gauge pressure. A multiscale gated recurrent unit (GRU) network estimates contact force from angular histories to account for the the combined effects of geometric deformation and material response. At zero gauge pressure, the model achieves a test RMSE of 0.76~N. Positive gauge pressure approximately doubles the average axial loading stiffness at +20~kPa over the first 3~mm of compression, while pressure-specific models maintain RMSEs below 0.93~N. Under negative gauge pressure, an energy-based compensation model enables contact force estimation using the the model trained at zero gauge pressure, achieving RMSEs of 1.66--2.10~N under quasi-static conditions. These results demonstrate a compact approach to combining force feedback, tunable axial stiffness, and self-sensing actuation in soft origami systems.

cs.RO

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02139

Adversarial Robustness of Retrieval-Based Fact-Checking: Detecting and Mitigating Evidence Contamination

Alamgir Munir Qazi, John P. McCrae, Jamal Abdul Nasir

原文摘要

Retrieval-based fact-checking systems ground verification in large document collections, creating a critical vulnerability: adversarial evidence contamination. We demonstrate that factually inconsistent but contextually coherent evidence generated to satisfy both retrieval-similarity and linguistic-fluency constraints can systematically mislead verification models. Contaminating just 3.1-7.4% of training corpora with five label-specific strategies causes adversarial content to penetrate up to 35.2% of top-k retrievals, degrading verification accuracy by up to 46.9% relative F1. Existing retrieval-stage defenses, validated against gradient-optimized or syntactically anomalous attacks, transfer poorly to this threat model: detection rates collapse and F1 recovery exhibits high variance across seeds. We propose GLAD (Geometric-Lexical Adversarial Defense), which fuses dense embedding peer-coherence with lexical distributional isolation. Across two benchmarks, two embedding models, and five random seeds, GLAD achieves the highest detection rates (70-86%) and the most consistent F1 recovery, reclaiming 73-80% of contamination-induced degradation on LIAR-RAW while approaching clean baselines on RAWFC. Crucially, unlike defenses that trade clean performance for robustness, GLAD imposes negligible cost on uncontaminated corpora.

cs.IR · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2606.05134

Activation-Based Active Learning for In-Context Learning: Challenges and Insights

Yaseen M. Osman, Geoff V. Merrett, Stuart E. Middleton

原文摘要

Deep active learning has previously been explored for LLM in-context sample selection, but not with methods that utilise recent advances in understanding of transformer activations. In this paper, we test the hypothesis that model activations could provide a fine-grained signal to optimise the selection of in-context examples. We present a comprehensive analysis of MLP activation-based deep active learning methods applied to in-context learning, including how different attention masking strategies impact active learning across diverse classification and generative datasets, using both Llama-3.2-3B and Qwen2.5-3B base models. However, we find a negative result: MLP and embedding layer outputs, viewed through the lenses of massive activations or the first four moments, do not correlate with example quality or task performance. Specifically, the absolute Spearman correlation coefficient is at most 0.33 for all tasks and models we tested, showing that such activation-based sampling should not be used for in-context learning. We hypothesise that this may be due to superposition, whereby models represent more features than they have dimensionality, suggesting that methods like Sparse Autoencoders (SAEs) may be a promising future direction.

cs.CL · cs.LG

查看论文
arXiv ↗PDF ↗
新论文 · v1arXiv:2611.02134

HaloGuard 2.0: An Open-Weights Constitutional Output Guard for Multilingual Response Safety

Ashmiya Lenin, Navaneeth Sangameswaran, Preetham S

原文摘要

An input guard decides whether a request reaches the model, but never sees what the model says. Harm that emerges during generation, because a jailbreak passed the input layer or a benign request drifted into operational detail, is visible only on the output side and only as tokens arrive. We present HaloGuard 2.0, an open-weight streaming output guard that assigns a safe/unsafe probability to every response token in a single forward pass, so a deployment can stop the stream once harm is confirmed. The guard reuses the constitution of HaloGuard 1.0, our input guard, to organise a 478,532-response training corpus, and supervises each token from the point where the response becomes harmful. We find that a guard trained on conventionally shaped data learns a shortcut specific to streaming: it judges the prompt instead of the response, and handles only 13.9% of matched safe/unsafe response pairs correctly. We correct this with a second phase on 106,043 counterfactual responses built around the base guard's own failure modes, and release the equal-weight average of the base and corrective models. On six English response-safety benchmarks, HaloGuard 2.0-0.8B reaches a macro F1 of 83.6 (83.3 on held-out halves of the decision-rule search), up from 72.8, with 38% fewer false positives. It matches the strongest full-response guards with 5-10x fewer parameters and is 2.6 points above the strongest streaming guard, Qwen3Guard-Stream-4B. A multilingual variant, HaloGuard 2.0-Poly-0.8B, trained on PolyGuardMix, reaches 83.1 F1 on the response labels of PolyGuardPrompts across 17 languages, 5.0 points above the strongest baseline. On one H100 the guard costs 0.37 ms per token per stream when moderating 64 concurrent responses. Both guards are hardened continuously through the weekly red-teaming protocol of HaloGuard 1.0, extended here to the output layer.

cs.CL · cs.CR

查看论文
arXiv ↗PDF ↗