跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+1743新增论文
+1399其中 AI 相关
+191已有论文修订
+94新增 AI 导读

arXiv 最近完成扫描:2026-11-19 19:10 · 全站最近入库:2026-11-19 19:09各信源进度 →

统计口径与数据新鲜度

统计 2026-11-18 19:27 — 2026-11-19 19:27(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 19:10最近发现新增版本:2026-11-19 19:09库内论文最新更新时间:2026-11-19 02:59

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

27,982 已收录论文23,249 AI 相关8,559 有修订112 HF 社区发现

浏览论文库

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 20,663 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-19 19:07(北京时间)
版本更新 · v2arXiv:2410.18270

Multilingual Hallucination Gaps in Large Language Models

Cléa Chataigner, Afaf Taïk, Golnoosh Farnadi

原文摘要

Large language models (LLMs) are increasingly used as alternatives to traditional search engines given their capacity to generate text that resembles human language. However, this shift is concerning, as LLMs often generate hallucinations, misleading or false information that appears highly credible. In this study, we explore the phenomenon of hallucinations across multiple languages in freeform text generation, focusing on what we call multilingual hallucination gaps. These gaps reflect differences in the frequency of hallucinated answers depending on the prompt and language used. To quantify such hallucinations, we used the FactScore metric and extended its framework to a multilingual setting. We conducted experiments using LLMs from the LLaMA, Qwen, and Aya families, generating biographies in 19 languages and comparing the results to Wikipedia pages. Our results reveal variations in hallucination rates, especially between high and low resource languages, raising important questions about LLM multilingual performance and the challenges in evaluating hallucinations in multilingual freeform text generation.

cs.CL

新论文 · v1arXiv:2611.21394

Beyond Bilingual: Studying Quadrilingual Code-Switching with Conversational AI

Annika Bush, Melika Minnaar, Salys Sultan, Tasnimul Hasan Tashin, Ankur Bhatt, Meltem Aksoy

原文摘要

Multilingual speakers choose languages by topic and conversation partner. Large language models (LLMs) support these languages unequally. Research on code-switching with conversational AI has mostly examined bilingual pairs of well-resourced languages. In Curaçao and Aruba, many people use Dutch, English, Spanish, and the low-resource Creole Papiamentu/o. In a preregistered study, 42 university students held two five-minute chatbot conversations, one about university life and one about local food and culture. Surveys assessed language background, AI use, and switching comfort. 20 international students in Curaçao formed a comparison sample. Participants switched between turns and rarely mixed languages within a message. English dominated both conversations. Participants switched more often in the cultural conversation, where Papiamentu/o and Spanish rose from 15% to 24% of messages. Self-reports showed no lower comfort switching with the chatbot than with a person. Conversational agents must follow turn-by-turn language changes and handle local low-resource languages well.

cs.HC · cs.CL

新论文 · v1arXiv:2611.21393

Train After You Measure: a Pre-Registered Learning Curve for Fine-Tuning Typed Decision Models on One User's Labels

Federico Lucca

原文摘要

Fine-tuning a small typed decision model on one user's email labels did not clear the margin we pre-registered over a logistic regression trained on the same labels. In earlier work, about two hundred labels and a TF-IDF logistic regression had the highest point accuracy on this inbox, though not significantly above the best hosted model; we ask whether the same labels, given to a decision model, close the gap. A first registered study, on a generic 4B base with cross-validation and message type only, stayed below the baseline at every label budget (-1.8 to -7.3 points). The main study was registered before any fine-tuning run: a fixed test set of 135 emails never used for selection, LoRA adapters on the public Tev1-0.8B and Tev1-4B decision models, nested subsets of 50 to 587 labels, two seeds and five decision fields on a free T4 GPU. Four of its five hypotheses are refuted--two in the fine-tunes' favour--and one is confirmed for one model. On message type the 4B model misses the registered three-point margin at 400 labels (+2.6) and passes it with all 587 (+3.3); with all labels both models are above the baseline by +0.7 and +3.3 points, where one email is 0.74 points and a 95% interval is about 15 points wide, and the 4B model is clearly ahead in type macro-F1 (0.581 vs. 0.406). On the three action switches the baseline is ahead at every label budget for both models (0.931 vs. 0.904 and 0.899 mean accuracy with all labels), at under a millisecond per question against 306 and 1681 ms. The 0.8B type curve barely improves after 300 labels and then falls. The baseline never predicts the two rarest classes (7 offers and 8 spam in the training pool); the fine-tuned models find spam in some runs, not consistently across seeds. We list the explanations we did not test and say when a new attempt would be worth making. Scripts are public; weights and emails are not.

cs.CL · cs.LG

新论文 · v1arXiv:2611.21392

VideoNexus: Video Perception Beyond the Reasoning Context

Yancheng Long, Xuanyu Zheng, Wei Chen, Luoyuan Zhang, Haoyang Xu, Bin Wen, Tingting Gao

原文摘要

Understanding long videos requires locating evidence, resolving visual details, and integrating observations into an answer. Even when relevant evidence is present, performing these operations within long, dense visual sequences burdens both perception and reasoning. Video agents enable selective inspection, but returning frames to the planner still accumulates visual information in its reasoning context. Delegating perception reduces this burden, yet sparse sampling can miss decisive details, and per-frame descriptions can preserve context growth. We introduce VideoNexus, a training-free framework that separates detailed video perception from evidence-driven planning and reasoning. The planner controls where to inspect, what to resolve, and how densely to sample. Perception sub-agents process native audio-video in independent contexts and return compact, temporally grounded evidence for iterative search and verification. Using Gemini 3.8 Flash for both planning and perception, VideoNexus achieves 88.6% accuracy on full LVBench, exceeding same-backbone static ingestion and built-in agentic processing by 3.6 and 1.1 points, respectively, at 44% of the latter's measured API cost. A paired LVBench subset ablation shows lower planner input and higher accuracy. With Kimi K3 as the shared backbone, VideoNexus improves over the official Video-GAIA ReAct harness by 8.5 points in our evaluation. These results support compact evidence exchange for combining detailed observation with efficient planning and reasoning.

cs.CV

新论文 · v1arXiv:2611.21389

Vanishing Relative Distance of Uniformly Random CSS LDPC Codes Built from Permutation Matrices

Kenta Kasai

原文摘要

Classical random regular low-density parity-check codes can have minimum distance growing linearly with block length. We study CSS check matrices formed as full $J\times L$ arrays of arbitrary $P\times P$ permutation matrices, sampled uniformly subject to commutation. For fixed column weight $J\ge3$ and fixed even row weight $L>2J$, we prove that the relative minimum distance converges to zero in probability as $P\to\infty$, with block length $n=LP$. We analyze the weight enumerators that count all nontrivial logical representatives after excluding stabilizers. A coefficientwise lower bound shows that, at every fixed relative weight $0<α<1$ along even-weight sequences, exponentially many such representatives exist with high probability. Their typical and mean counts have exponent lower bound $h_2(α)/2$, where $h_2$ is the binary entropy function. We compare this bound with the classical LDPC and dense CSS mean weight-distribution exponents. The existence, with high probability, of a logical operator below every fixed target weight $δn$ yields the distance conclusion.

quant-ph · cs.IT

版本更新 · v2arXiv:2608.00111

FDIR: Fidelity-Anchored Decompressed Image Restoration for Human and Machine Perception

Kuan-Yen Chen, Chung-Yao Ma, Fang-Yi Su, Philip Chikontwe, Jonathan Guo, Pu-Jen Cheng, Jung-Hsien Chiang

原文摘要

Evaluating restored images requires considering pixel fidelity, human perception, and machine preference. However, existing compressed image restoration methods rarely satisfy all three criteria at once. Regression models preserve pixel fidelity but often produce overly smooth results. Conversely, generative models create realistic textures but tend to drift from the ground truth. To address this trade-off, we propose FDIR (Fidelity-Anchored Decompressed Image Restoration), a framework that uses generative priors while preserving strong pixel fidelity. FDIR works in two stages. First, a quality-guided one-step flow model (QO-Flow) adapts a pretrained latent flow backbone with LoRA to generate a structural prior in a single forward pass. Second, rather than outputting this generated image directly, a flow-conditioned detail refinement network (FCDR) uses it to predict a pixel-space residual on the decompressed input. This design brings in perceptual details while keeping the final output faithful to the original image. We evaluate FDIR on LIVE1 and DIV2K under severe compression at quality factors 1 and 5. Across all four settings, FDIR achieves the highest PSNR among the seven compared methods and beats the strongest perceptual baseline, CODiff, by 1.3 to 2.1 dB. It also ranks second on LPIPS in all four settings and on FID and DISTS on DIV2K-Val, right behind CODiff. For downstream detection, segmentation, and retrieval at quality factors 20 and 30, FDIR ranks second on every individual metric and achieves the best average rank across all ten metrics. Overall, FDIR is the only method to maintain a top-two average rank across fidelity, perception, and machine preference.

eess.IV · cs.CV · cs.LG

版本更新 · v2arXiv:2607.16927

Deep Adaptive Bayesian Screening

Jade Lejeune Herman, Arno Strouwen, Johan A. K. Suykens, Peter Goos

原文摘要

We introduce Deep Adaptive Bayesian Screening (DABS), a method for performing adaptive factorial screening in discrete design spaces under tight experimental budgets. DABS learns a policy network offline to sequentially select informative experiments, amortizing Bayesian Optimal Experimental Design. It handles binary designs, incorporates sparsity and interactions via a spike-and-slab prior with strong heredity. The model is trained using a contrastive lower bound on information about factor activity with nuisance effect sizes and noise variance analytically integrated out. Unlike prior amortized Bayesian design approaches, DABS also integrates posterior inference at deployment, yielding posterior probabilities of factor activity and credible intervals on effect sizes. We evaluate DABS on simulated screening problems calibrated to published experimental benchmarks, and show that it matches or outperforms the strongest non-adaptive Bayesian baseline, with the largest gains when the budget is small relative to the number of candidate factors.

stat.ML · cs.LG

新论文 · v1arXiv:2611.21384

Stability-Compensated Threshold Calibration for Selection-Conditioned Risk Control in Large Language Models

Zhiyuan Wang, Yuanchang Ye, Jiahao Yuan, Jinhao Duan, Xiaoshuang Shi, Weijun Xie, Ye Wei, Kaidi Xu

原文摘要

Large language models often produce hallucinated answers, motivating selective prediction policies that abstain on uncertain queries. However, heuristic uncertainty thresholds alone do not provide statistical control of errors among accepted predictions. Linear Expectation Constraints (LEC) reduce selection-conditioned error rate (SCER) calibration to a nonpositive expected linear-loss constraint, but the associated loss is non-monotonic in the acceptance threshold. We construct an exact counterexample under i.i.d. sampling demonstrating that LEC's empirical threshold-selection rule can violate the population SCER target. To address this risk-transfer gap, we introduce Stability-Compensated Linear Expectation Control (SC-LEC), which uses the negative linear risk of a theoretical full-data reference to compensate for threshold-selection instability. For local expansion beyond a safe anchor, we derive two finite-sample certificates: a stability-based certificate exploiting rank crossings and a direct certificate using the structure of threshold expansion to reduce conservatism. With a valid population anchor margin, our certificate conditions ensure finite-sample marginal SCER control, and independent margin certification further provides a $1-δ$ high-probability guarantee for this calibration-averaged control. Experiments on CommonsenseQA, ARC-Challenge, and TriviaQA show useful retention gains beyond the same certified anchor, while revealing conservative behavior in small-margin regimes.

cs.CL

新论文 · v1arXiv:2611.21383

The Mask Is Not the Object: Volumetric Supervision for 3D Gaussian Segmentation

Jeonghwan Cho, Minsu Kim, Junyoung Hong, Seon Joo Kim

原文摘要

3D Gaussian Splatting enables object-level operations on Gaussian primitives, yet existing segmentation methods are largely trained and evaluated through rendered views. This mismatch can produce accurate rendered masks from incomplete 3D selections since alpha-blended feature supervision concentrates learning on dominant contributors while leaving occluded or low-contribution Gaussians weakly supervised. We propose VODA, a plug-and-play supervision objective that exploits local depth structure to group per-pixel contributors and directly supervises all Gaussians in selected depth clusters, bypassing contribution-weighted gradient delivery. To evaluate volumetric completeness, we introduce Removal-3D, a synthetic object-removal benchmark, and the Removal Depth Error (RDE), a metric comparing predicted and ground-truth depth changes. Across four rendering-based baselines, VODA consistently improves volumetric segmentation on Removal-3D while maintaining or improving performance on LERF and ScanNet. Further analysis shows that VODA recovers lower-contribution Gaussians, while simple gradient reweighting and spatial propagation fail to reproduce comparable gains.

cs.CV · cs.AI

版本更新 · v2arXiv:2610.40749

Combinatorial Residual Network: a decay-chain-aware classifier without explicit jet assignment

John Alison, Patrick Bryant, Alejandro Gomez Espinosa, Christian Guinto-Brody, Aniket Khanal, Chuyuan Liu, Sindhu Murthy, Guanglei Zhao

原文摘要

Final states with several jets from the decays of heavy particles pose a combinatorial problem: the jets must be assigned, at least implicitly, to their parent particles. The Combinatorial Residual Network (CoReNet) is a classifier that encodes the assumed decay chain in its architecture, building the event representation level by level along the decay hierarchy and combining all jet assignment hypotheses, evaluated with shared weights, with learned weights. The exchange symmetries of the decay are built in, the network size does not depend on the number of hypotheses, and no generator-level jet assignments are used in training. Developed for the $b\bar{b}b\bar{b}$ final state and used in several CMS searches, CoReNet with 1508 parameters performs well above kinematic baselines on fast-simulated Higgs boson pair production against QCD multijet events and trains in minutes on a single GPU, and its hypothesis weights select the truth-matched jet pairing in most signal events in which it is defined. Since only the encoded decay chain is specific to the final state, we test the construction on $HH \to b\bar{b}W^+W^-$ against semileptonic $t\bar{t}$, where encoding the decay chain of the background as well improves the classification at a cost of about $10^{3}$ parameters. In both final states, CoReNet performs comparably to a state-of-the-art transformer-based network more than two orders of magnitude larger. The implementation is publicly available at https://github.com/cms-cmu/CoReNet.

hep-ex · cs.LG · hep-ph · physics.data-an

新论文 · v1arXiv:2611.21379

What's the Difference? Voice Design and Confirmation Bias in AI-Assisted Self-Diagnosis: Anthropomorphic vs. Machine-Like Voice Output

Niklas Roethlingshoefer, Emma Winkler, Melanie Loeffler, Emely Rosbach

原文摘要

As symptom checkers based on artificial intelligence (AI) are becoming common entry points for health information, their influence on lay users' diagnostic decisions raises questions about trust and bias. Of particular concern is confirmation bias, the tendency to favour information that supports one's existing belief while discounting information that contradicts it. One of the central design choices for such systems is the style of voice: whether they should present themselves with a human-like (anthropomorphic) or a clearly machine-like (xenomorphic) voice. This study investigates whether the voice design of an AI symptom checker modulates confirmation bias in self-diagnosis. In a within-subjects online experiment (N = 81 participants, 648 trials), participants selected a diagnosis and rated their confidence, then heard an AI-generated recommendation that either confirmed or contradicted their initial choice, delivered by an anthropomorphic or a machine-like voice, before deciding again. We observed a strong, direction-dependent influence of the AI: participants revised their diagnosis in 45.7% of trials when the AI contradicted them versus 1.2% when it confirmed them (generalised linear mixed model, GLMM, p < .001); confidence rose under confirmation but barely moved under contradiction (robust linear mixed model, p < .001); and decisions took longer after contradiction (Gamma GLMM, p < .001). This effect did not differ between the two voices (no voice or interaction term was significant, smallest p = .282), even though a manipulation check confirmed the voices were perceived as markedly different. The findings suggest that the direction of an AI recommendation, rather than the anthropomorphism of its voice, drives confirmation-related bias in AI-assisted self-diagnosis.

cs.HC

版本更新 · v2arXiv:2607.07386

Sparse Delta Memory: Scaling the State of Linear RNNs through Sparsity

Loïc Cabannes, Pierre-Emmanuel Mazaré, Ilze Amanda Auzina, Théo Le Pendeven, Gergely Szilvasy, Matthijs Douze, Maria Lomeli, Justin Carpentier 等 10 位作者

原文摘要

Linear attention models allow a fixed state size and a fixed amount of compute per token. However, due to their limited state size, linear attention models fall behind in long-context recall compared to softmax-attention-based transformer architectures. Increasing the state size of linear attention improves recall performance but at the cost of higher FLOPs. In this work, we introduce Sparse Delta Memory (SDM), an architecture that scales the hidden state of gated linear RNNs to orders of magnitude higher capacity using a sparse addressing scheme. SDM extends the Gated DeltaNet architecture by replacing the dense key-value outer product with sparse reads and writes to a large explicit memory. We show that, under an isoFLOP constraint and with an identical number of parameters, a higher state memory capacity significantly improves performance on in-context learning and long-context retrieval tasks. Moreover, by learning the initial state of the SDM memory and therefore using it as a parametric memory, we show that the model further improves on a wide range of common-knowledge and reasoning tasks.

cs.LG

新论文 · v1arXiv:2611.21378

Data-Driven Predictive Control for Autonomous Vehicle Racing under Varying Friction

Julius Beerwerth, Julius Tischbein, Bassam Alrifaee

原文摘要

Data-driven predictive control (DPC) predicts future outputs by combining recorded input-output trajectory data such that the combined trajectory approximately matches a measured initial trajectory and tracks a provided reference. This paper investigates to what extent and under which conditions DPC using data pooled across operating conditions retains the prediction accuracy of DPC using data from the current operating condition. In autonomous vehicle racing, the tire-road friction varies between racetrack sections and tire sets. If the initial trajectory and the recorded data originate from different frictions, prediction accuracy of DPC degrades. Ensuring the frictions match requires knowing the friction during recording and estimating the current friction explicitly. Instead, we pool data recorded at different frictions. We compare the prediction accuracy of DPC using data recorded under the current friction (matched), a different friction (mismatched) and data pooled from both frictions. We evaluate in simulation and small-scale vehicle experiments in the Cyber-Physical Mobility Lab (CPM Lab). Our results show that pooled data keep the multi-step root-mean-square prediction error within 1.5 times that of matched data, whereas mismatched data increase it by a factor of 2.2 to 3.4. To assess how this prediction accuracy carries over to control, we evaluate the tracking error from the raceline in closed loop. In simulation, DPC with pooled data tracks comparably to nonlinear model predictive control (13.7 cm versus 13.3 cm) and close to DPC with matched data in the CPM Lab, differing by 0.1 cm on the low-friction tires (1.6 cm versus 1.5 cm) and by 0.2 cm on the high-friction tires (0.9 cm versus 1.1 cm). DPC thus provides partial adaptation to varying friction, without a dedicated adaptation method.

eess.SY · cs.RO

新论文 · v1arXiv:2611.21372

Learning What Pairs Miss for Higher-Order Multimodal Alignment

Kunyu Zhang, Shenghui Feng, Xiangjie Shi, Zeqing Wang, Yingxu Wang

原文摘要

Contrastive learning beyond two modalities now scores the whole tuple with multilinear products, volumes, and fusion heads, critics that can represent the joint distribution. In practice these critics still behave like bags of pairs. Their negatives break at least two of the three pairs, a single pairwise check rejects them, and when most samples can be solved by pairs the higher-order term receives almost no gradient. To solve this problem, we propose CHORD (Contrastive Higher-Order Residuals with Decoys), which trains a residual against decoys drawn from a calibrated pairwise model, negatives that the pairwise model finds plausible, so that rejecting them often needs all three modalities. Drawing decoys costs one inner-product query over a memory bank, and the optimal residual is the log-ratio between the data and the pairwise model, so the decoy loss also bounds what that model misses. On M3-Mix, where one in ten training samples needs all three modalities, our CHORD improves accuracy on those samples by 21.2 percentage points over the strongest of eight baselines. CHORD also gives the best two-to-one results on every real dataset, and its residual improves other retrievers without retraining.

cs.LG

新论文 · v1arXiv:2611.21370

Federated TD Learning in Heterogeneous Environments with Average Rewards: A Two-timescale Approach with Polyak-Ruppert Averaging

Ankur Naskar, Gugan Thoppe, Abbasali Koochakzadeh, Vijay Gupta

原文摘要

Federated Reinforcement Learning (FRL) provides a promising way to speedup training in reinforcement learning when one has access to multiple edge devices that can operate in parallel. Recently, it has been shown that even when these edge devices have access to different dynamic models---such as cars or robots with differing system configurations---an optimal convergence rate that has a linear speedup proportional to the number of devices is achievable. However, this result requires that the algorithm's stepsize be chosen in a manner that depends on the unknown model parameters. Also, it applies only to a discounted setting, which has been argued to fit episodic tasks better than continuing control tasks. In this paper, we obtain finite-time bounds for heterogeneous FRL with average rewards. We show that optimal convergence rate and linear speedup are achievable even with a universal stepsize choice, independent of the underlying dynamics. To achieve our result, we modify the existing one-timescale FRL method to a novel two-timescale variant that additionally incorporates iterate averaging. We remark that this idea also extends to other settings already studied in the literature, such as the homogeneous FRL or the heterogeneous case with a discounted cost function.

cs.LG

新论文 · v1arXiv:2611.21367

Premise Recovery in Software Design: What Does This Decision Rest On?

Priyabanta Sandulua, Islam Mesabaha, Sebastian J. Vollmer

原文摘要

Design decisions shape software architecture, and writing them down preserves the reasons behind it. Architecture decision records (ADRs) capture them, each with its context, choice, and consequences. A decision can depend on earlier ones: reversing one of them would force its review. We call these its premises. Yet ADRs do not always name their premises. When a decision is reversed, engineers supersede its ADR and search by hand for what it affects. That search can miss ADRs resting on an unnamed premise. We frame this as premise recovery: given an ADR with identifiers masked, find its premises among earlier records. We collect 1,060 ADRs from 36 repositories and label 1,529 candidate pairs for 60 target ADRs. Two LLM annotators agree at \k{appa} = 0.859. An independent human annotator, blind to their labels, agrees with the stated/implicit tag on 30 of 32 premises (\k{appa} = 0.875). We compare 7 open-weight models with 2 baselines that use no LLM. Qwen3.8-27B reaches F1 0.756, against 0.466 for word overlap; 2 models added post hoc score similarly. On the 11 premises whose titles barely overlap the target, it finds 0.758, while word overlap finds none. Reasoning raises its recall of implicit premises but not its F1. Word overlap with as many predictions gains 69% as much. We find only 2 to 3 stale dependents, though we searched for unlinked ones in only 30 of 66 supersessions. Open models could thus suggest unwritten premises when a decision is reversed, with precision between 0.68 and 0.78.

cs.SE · cs.AI · cs.MA

新论文 · v1arXiv:2611.21362

What You Post Is What You Symptom: Estimating Psychometric Networks from Social Media Discourse

Ştefana-Arina Tăbuşcă, Ana Sabina Uban, Liviu P. Dinu

原文摘要

Symptom networks in psychopathology are estimated almost exclusively from self-report questionnaires, restricting inference to whatever symptom set a given instrument covers. We introduce a framework for symptom-resolved network estimation directly from social media discourse and apply the framework to depression and eating disorders, with a third anorexia corpus as cross-dataset validation. To our knowledge, we are the first to estimate symptom-resolved psychometric networks from social media text, compare them to clinical co-endorsement networks, and replicate transdiagnostic linguistic markers across 3 independently recruited corpora spanning 2 disorders simultaneously. For depression, expression scores correlate with BDI-II totals at $ρ{=}0.434$, 20 of 21 items reach FDR significance, and split-half reliability is 0.935. The Network Comparison Test detects few edge differences between modalities (3-5\% of edges), while bridge symptoms diverge. For eating disorders, the item-level results are exploratory. Identified linguistic features replicate the same direction across all 3 corpora, and they cluster in three construct families.

cs.CL · cs.AI · cs.CY · cs.SI

版本更新 · v3arXiv:2605.17703

EduSOC: A Lightweight Simulated Security Operations Center with AI Incident Responders

Martin Higgins, Shawn Thompson, Cherry Mangla, Bobby Swar

原文摘要

This paper presents EduSOC, a lightweight web-based Security Operations Center (SOC) simulation platform for cybersecurity education. Conventional cyber ranges can be complex and resource-intensive to deploy, particularly for introductory exercises focused on investigative decision-making rather than infrastructure configuration. EduSOC provides an alternative by combining continuous synthetic event generation, searchable supporting evidence, and role-based artificial intelligence (AI) responders that assist investigations without revealing underlying incident classifications. Students remain responsible for interpreting evidence, consulting simulated operational teams, and justifying their escalation decisions. A preliminary classroom evaluation was conducted across three university cohorts, comprising 50 matched pre- and post-simulation survey participants. The findings indicated high student engagement and improvements in self-reported SOC self-efficacy, while short-term knowledge gains varied across cohorts. These findings provide empirical insight into the use of lightweight, AI-supported SOC simulations in university cybersecurity education.

cs.CR · cs.HC

版本更新 · v2arXiv:2603.17896

A Noise Sensitivity Exponent Controls Large Statistical-to-Computational Gaps in Single- and Multi-Index Models

Leonardo Defilippis, Florent Krzakala, Bruno Loureiro, Antoine Maillard

原文摘要

Understanding when learning is statistically possible yet computationally hard is a central challenge in high-dimensional statistics. In this work, we investigate this question in the context of single- and multi-index models, classes of functions widely studied as benchmarks to probe the ability of machine learning methods to discover features in high-dimensional data. Our main contribution is to show that a Noise Sensitivity Exponent (NSE) - a simple quantity determined by the activation function - governs the existence and magnitude of statistical-to-computational gaps within a broad regime of these models. We first establish that, in single-index models with large additive noise, the onset of a computational bottleneck is fully characterized by the NSE. We then demonstrate that the same exponent controls a statistical-computational gap in the specialization transition of large separable multi-index models, where individual components become learnable. Finally, in hierarchical multi-index models, we show that the NSE governs the optimal computational rate in which different directions are sequentially learned. Taken together, our results identify the NSE as a unifying property linking noise robustness, computational hardness, and feature specialization in high-dimensional learning.

stat.ML · cs.LG

版本更新 · v6arXiv:2512.00324

MILE: A Mechanically Isomorphic Hand Exoskeleton and Visuotactile Robotic Hand for Data Collection in Dexterous Manipulation

Jinda Du, Jieji Ren, Qiaojun Yu, Ningbin Zhang, Yu Deng, Xingyu Wei, Yufei Liu, Guoying Gu 等 9 位作者

原文摘要

Dexterous robotic hands perform complex, contact-rich manipulation. Imitation learning provides a route to such skills, but collecting human demonstrations with accurate hand actions and rich tactile information remains a key bottleneck. We present MILE, a teleoperation-based data-collection system comprising the wearable MILE exoskeleton and the mechanically corresponding MILE-Tac robotic hand. Because human-hand anatomy and wearability place tighter constraints on the high-DoF wearable, our human-first design begins with the MILE exoskeleton, equipped with custom modular joint encoders for accurate joint-angle acquisition. We then design the MILE-Tac robotic hand to share the exoskeleton's selected kinematic topology and joint-axis arrangement while satisfying robot-side implementation constraints, and equip its fingertips with compact visuotactile sensor modules. This correspondence enables direct exoskeleton-to-robot joint-space command transfer without online task-space inverse-kinematics retargeting. During teleoperation, the system synchronously records task-specific visual observations, four fingertip visuotactile streams, robot-hand proprioception, and exoskeleton-derived action commands. In a four-task teleoperation benchmark, MILE achieved a mean success rate of 76%, compared with 28% and 8% for glove-based and vision-based baselines, respectively. For downstream imitation learning, we trained paired ACT and DP policies with and without tactile input on MILE-collected demonstrations. The tactile-input variants achieved higher success rates in all paired evaluations.

cs.RO · cs.CV · cs.HC

版本更新 · v3arXiv:2606.16583

Uncertainty Is Not a Safety Net for Clinical VQA, but Can It Anticipate Model Failure?

Arnisa Fazla, Alberto Testoni, Ameen Abu-Hanna, Barbara Plank, Iacer Calixto

原文摘要

Safe deployment of clinical vision-language models (VLMs) requires reliable uncertainty estimation (UE): a signal indicating when predictions should be trusted or escalated to a clinician. We test whether current UE methods actually deliver this signal. Benchmarking 8 methods across 12 VLMs on clinical visual question-answering (VQA), we find that UE quality is not an intrinsic property of the UE method: it tracks model accuracy, degrading precisely where the model performance is weakest, and therefore where reliability is most needed. When we stress-test models by hiding the correct option among the multiple-choice answers (NOTA perturbations), accuracy collapses while uncertainty barely changes, leaving models systematically miscalibrated. Yet, we find that uncertainty on the unperturbed input reliably anticipates which predictions will collapse under NOTA, indicating that UE in current VLMs carries diagnostic information about model fragility. Our results position UE as a diagnostic tool for identifying fragile predictions and motivate perturbation-based evaluation as a path toward safe clinical deployment.

cs.CL

新论文 · v1arXiv:2611.21352

Specificity-Guided Knowledge Diversion for Decomposable Generative Models

Yucheng Xie, Fu Feng, Jing Wang, Xin Geng, Yong Rui

原文摘要

Generative model pre-training typically optimizes a model as a whole without explicitly organizing its parameters for selective reuse. Decomposable pre-training instead learns reusable components for downstream adaptation while retaining strong generative performance. However, existing knowledge diversion methods incur substantial parameter overhead through weight factorization and rely on fixed component assignments to induce specialization, potentially compromising in-distribution generation quality. We therefore propose Specificity-Guided Knowledge Diversion (SGKDiv), which learns how generative knowledge is shared and specialized for effective generation and selective downstream reuse. SGKDiv constructs Knowledge Units from paired feed-forward channels, preserving native computational structure without SVD-based reparameterization. Calibrated condition-wise routing combines a shared usage prior with conditional adjustments, while complementary regularization shapes their allocation, diversity, and magnitude. Together, these designs replace predefined component assignments with adaptively learned and explicitly regulated knowledge specialization. Experiments across diverse model and training settings demonstrate better in-distribution generation than dense and MoE baselines, alongside stronger downstream adaptation than representative PEFT methods across eight datasets.

cs.LG · cs.CV

版本更新 · v2arXiv:2611.06207

Performance Portability of Out-of-Kernel Optimizations on GPUs: A Molecular Docking Case Study

Gianmarco Accordi, Leonardo Beltrame, Theresa Pollinger, Jens Domke, Davide Gadioli, Gianluca Palermo

原文摘要

Achieving performance portability is challenging for irregular scientific applications built as repeated multi-kernel GPU pipelines, because throughput depends on out-of-kernel decisions as much as on kernel efficiency. Using molecular docking as a representative case study, we propose an out-of-kernel tuning methodology that treats batching policy, worker concurrency, and per-worker memory allocation as explicit optimization parameters, replacing backend-dependent occupancy-driven choices with a more portable execution strategy based on explicit resource bounds. We implement this methodology in the muDock mini-app under a manager/worker execution model where multiple host workers share a GPU through independent streams. We then evaluate the methodology on NVIDIA, AMD, and Intel GPUs across CUDA, HIP, and SYCL implementations. Across all evaluated platform/backend pairs, the memory-bounded formulation exceeds the peak throughput of the occupancy-guided baseline. The fully tuned methodology improves the end-to-end throughput on all tested configurations, achieving an average speedup of 1.53x over the baseline.

cs.DC · cs.PF

新论文 · v1arXiv:2611.21348

Representational Change Under Self-Consuming Training: A Sparse Autoencoder Analysis

Suchit Gupte, Zhongteng Cai, Xueru Zhang, Mohammad Mahdi Khalili

原文摘要

Large language models are increasingly exposed to model-generated text, raising the possibility that future models will be trained recursively on the outputs of their predecessors. Prior work on such self-consuming training loops has largely focused on output quality and downstream performance, leaving their effects on internal representations poorly understood. We study how representations evolve across five generations of self-consumption and whether sparse autoencoders (SAEs) remain reliable under this shift. At each generation, we compare a lineage trained on synthetic data with a matched control trained on real data under the same training settings. We evaluate how well SAEs trained before self-consumption transfer to later generations, measuring reconstruction and sparsity across SAE widths, architectures, and model layers. We then fine-tune the original SAE on each generation and use sparse probing to test whether task-relevant information remains concentrated in a small number of latents or becomes more distributed across the representation. We find that zero-shot SAE performance degrades progressively under self-consumption, with growing reconstruction and sparsity gaps that vary substantially across layers. Fine-tuning largely restores reconstruction, but not feature-level usefulness: task-relevant information remains accessible in the full latent space while becoming less concentrated in individual features. These results show that self-consumption induces representational drift and that reconstruction fidelity alone is insufficient to establish that an adapted SAE remains equally useful for interpretation.

cs.LG · cs.AI

版本更新 · v3arXiv:2602.02190

PCA of probability measures: Sparse and Dense sampling regimes

Gachon Erell, Jérémie Bigot, Elsa Cazelles

原文摘要

A common approach to perform PCA on probability measures is to embed them into a Hilbert space where standard functional PCA techniques apply. While convergence rates for estimating the embedding of a single measure from $m$ samples are well understood, the literature has not addressed the setting involving multiple measures. In this paper, we study PCA in a double asymptotic regime where $n$ probability measures are observed, each through $m$ samples. We derive convergence rates of the form $n^{-1/2} + m^{-α}$ for the empirical covariance operator and the PCA excess risk, where $α>0$ depends on the chosen embedding. This characterizes the relationship between the number $n$ of measures and the number $m$ of samples per measure, revealing a sparse (small $m$) to dense (large $m$) transition in the convergence behavior. Moreover, we prove that the dense-regime rate is minimax optimal for the empirical covariance error. Our numerical experiments validate these theoretical rates and demonstrate that appropriate subsampling preserves PCA accuracy while reducing computational cost.

stat.ML · cs.LG

新论文 · v1arXiv:2611.21344

TEMPO: Regime-Aware Operator Learning with Locally Adapted POD Bases

Rodion Akinzhala, Alexander Terentyev, Alexey Golubev, Yury Maximov

原文摘要

Operator learning approximates the solution map of a parametric partial differential equation from data, enabling fast prediction at new parameter values without repeated simulation. The dominant approach fits a single global low-rank basis, typically a proper orthogonal decomposition (POD), to all solution snapshots. This fails when the snapshots span qualitatively different solution shapes, from smooth diffusive profiles to near-discontinuous shocks. We propose TEMPO (Trajectory Expectation-Maximization Mixture of POD Operators), a two-phase framework for regime-aware operator learning. In the offline phase, TEMPO recovers latent regimes via expectation-maximization, fitting a dedicated linear or nonlinear basis to each discovered regime. In the online phase, a gating network and per-regime branch networks are trained jointly, routing new inputs to the regimes at inference. We test TEMPO on Burgers' flow, Sod shock tubes, Coanda flow, Darcy flow, Navier-Stokes flow and the heterogeneous Helmholtz equation. With 64 total POD modes, TEMPO gives 18.61% test error on Burgers', compared with 19.61% for global POD. It also gives lower errors on Coanda and Helmholtz.

cs.LG

新论文 · v1arXiv:2611.21339

MovieRoll: A Systematic Framework for Rolling Out Long-Form Videos

Yunyang Ge, Weijie Kong, Ryan Xu, Yue Wu, Qi Tian, Jiangfeng Xiong, Bin Lin, Mingzhe Zheng 等 12 位作者

原文摘要

Streaming video generation must sustain quality beyond the training horizon while handling self-generated history and growing cache costs. To address these challenges, we present MovieRoll, a base-to-distilled framework that connects streaming fine-tuning, long-rollout distillation, and long-horizon inference. We first establish a robust multi-step base through Window-wise Teacher Forcing (WinTF), which jointly denoises chunks at progressive noise levels with bidirectional attention within each window and causal access to preceding history. To retain historical context efficiently, hierarchical KV-cache compression (HieraKV) exploits layer-wise attention patterns, retaining full history in shallow layers and sink--recent history in deeper layers. This reduces cache memory by up to 78% while improving generation quality. Building on this base, cumulative long tuning with distribution matching distillation (CuLT-DMD) holds parameters fixed throughout each long self-rollout and averages local gradients before updating. Each update therefore combines supervision across temporal positions while retaining only one local computation graph, enabling efficient long-rollout training of a few-step generator. To extend generation beyond the training horizon, KV recomputation bounds context for both models, while guidance annealing (AnnealCFG) reduces guidance toward the clean endpoint to stabilize MovieRoll-Base. Complementing these modeling advances, FBCA attention and PS-VAE decoding support efficient training and streaming output. On HeliosBench, MovieRoll-Base remains competitive at 30s after training on videos of up to 15s, and MovieRoll-Distilled achieves the highest total scores among the compared methods across single- and multi-event evaluations up to 60s, combining sustained naturalness with active motion.

cs.CV

新论文 · v1arXiv:2611.21335

Anchoring Motion, Refining Trajectories: Generalizable Continuous Neural Decoding

Fanqi Shen, Junru Chen, Surui Cheng, Jiahe Li, ZhiZhang Yuan, Gaorui Zhang, ZhiYuan Cao, Xiaoquan Shen 等 9 位作者

原文摘要

Brain activity encodes motor information across multiple temporal scales, enabling continuous neural control beyond discrete action recognition. However, existing methods are typically tailored to specific subjects and tasks, with direct neural-to-kinematic mappings that can entangle subject-specific neural variability with task-dependent motion patterns, limiting generalization. To address these limitations, we introduce AMRT, a shared neural decoding framework for heterogeneous electrode configurations and movement paradigms. Motivated by cortical encoding of movement direction and speed, AMRT adopts a two-stage design that anchors local motion through velocity estimation and refines the resulting trajectories over longer horizons, encouraging reuse of local motion tendencies across subjects and tasks. Across three datasets, AMRT generalizes to unseen trials and subjects and demonstrates transfer potential across neural recordings and movement paradigms, highlighting the potential of shared neural representations and anchor-and-refine decoding for generalizable continuous neural control.

q-bio.NC · cs.LG

新论文 · v1arXiv:2611.21334

Real-world Polarimetric Environment Map Dataset

Yonghee Oh, Ryota Maeda, Yunseong Moon, Seung-Hwan Baek

原文摘要

Environment maps are a standard representation of illumination, enabling realistic image-based lighting in modern rendering systems. However, existing environment maps typically encode only RGB radiance, omitting polarization, an intrinsic property of light that is imperceptible to the human eye but informative about surface shape and material properties. This omission precludes the physically accurate simulation of polarized light transport under in-the-wild illumination conditions, heavily limiting the development and evaluation of robust polarization-based scene reconstruction and understanding methods. We address this gap by introducing a real-world polarimetric environment map dataset, each providing a full-sphere panorama with per-pixel high-dynamic-range RGB Stokes vectors. To acquire the dataset, we design and calibrate a dedicated imaging system and develop a polarization-aware panorama-stitching pipeline. In particular, we introduce a frame-field representation for perspective cameras, together with singularity-aware blending, to enable smooth polarimetric panorama stitching without singularity artifacts. Using this system, we capture 100 polarimetric environment maps across diverse real-world scenes. We further develop a custom rendering plugin and demonstrate close agreement between real and rendered observations. Finally, we analyze the polarization state of the sky across different times and weather conditions, and demonstrate its practical impact on a downstream shape-from-polarization task.

cs.GR

新论文 · v1arXiv:2611.21331

TRIAGE-CPG: A Cross-Model Study of Budgeted Static Malware Triage with Large Language Models

Paul-Andrei Sava, Alexander Küchler, Daniel Kowatsch, Bernhard Grill, Jennifer Pullman, Yanick Fratantonio, Luca Invernizzi

原文摘要

When triaging unknown binaries, security analysts must decide which samples to clear as benign, flag as malicious, or escalate for deeper analysis. Large Language Models (LLMs) could support these decisions by interpreting decompiled code, but a binary can contain hundreds or thousands of functions, making exhaustive LLM analysis impractical. We study how current LLMs behave when static triage is formulated as budgeted evidence acquisition: which functions they inspect and how they trade off benign clearance, malware detection, and escalation. To enable a controlled comparison, we implement TRIAGE-CPG, a fixed pipeline over a code property graph (CPG) of a decompiled binary. An LLM-free heuristic selects seed functions, an agent expands the resulting evidence dossier under a fixed budget, and a reporting stage produces a benign, suspicious, or malicious verdict together with a behavioral report. We evaluate eight closed-source and open-weight models on 1,000 Windows PE binaries, using a 600-sample test split and the same zero-shot protocol for every model. We assess the ternary verdicts under both a strict binary reading and triage balanced accuracy (tBA), which credits malware that is flagged or escalated and benign samples only when cleared. After analyzing about 40 of a mean 1,000 functions per binary, the highest-scoring model reaches a strict balanced accuracy of 0.805 and a tBA of 0.855, flagging 65% of malware, escalating another 12%, and clearing 94% of benign samples. The models implement different triage policies: high-specificity models produce small, malware-rich escalation queues, whereas recall-oriented models escalate or flag much more broadly. Both closed-source and open-weight models appear across this spectrum. Our results identify no universal deployment configuration: model choice itself determines the clear/flag/escalate trade-off under limited static evidence.

cs.CR · cs.AI