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

+197新增论文
+151其中 AI 相关
+16已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-21 12:19 · 全站最近入库:2026-11-21 12:19各信源进度 →

统计口径与数据新鲜度

统计 2026-11-21 04:17 — 2026-11-22 04:17(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:2026-11-21 12:19最近发现新增版本:2026-11-21 12:19库内论文最新更新时间:2026-11-20 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。

29,158 已收录论文24,181 AI 相关8,992 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 21,490 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-21 12:17(北京时间)
版本更新 · v2arXiv:2610.02190

Trust the Direction, Search the Step: Zero-and-First-Order Methods for LLM Fine-Tuning

Cristian McGee, El Houcine Bergou, Aritra Dutta

原文摘要

Step-size selection remains a central challenge in large-scale neural network optimization; conservative steps slow convergence, while aggressive steps can destabilize it. We combine Zero-and-First-Order optimization (ZFO) and propose a lightweight framework that decouples direction selection from step-size. ZFO uses a trusted first-order optimizer to determine the direction and performs zeroth-order evaluations only along this one-dimensional subspace to choose how far to move. Using the current {gradient information} and two additional objective function evaluations, ZFO instances construct a local model of the objective function along the proposed direction and select a curvature-aware step within a bounded search interval. This yields an adaptive step-selection mechanism that costs less than a full line search. We provide theoretical guarantees to show that shared-sample evaluations produce reliable finite-difference curvature estimates, that the induced local model selects a near-optimal step along the search interval, and that ZFO converges to a neighborhood of a stationary point. Across the evaluated settings, language models and datasets, ZFO frequently improves optimization and final performance relative to fixed-step first-order baselines, with the magnitude and preferred local model depending on the objective. Our code is publicly available at: https://github.com/nizswan/Zeroth-First-Order-Framework.

cs.LG · math.OC

新论文 · v1arXiv:2611.21795

The Impact of Scenario Representation in Scenario-based Optimization Certification

Stephen Crawford, Nora Ayanian

原文摘要

Scenario-based optimization provides finite-sample guarantees for decisions constructed from sampled uncertainty, but its certification procedures can be sensitive to how those samples are represented. We investigate this sensitivity under lossless scenario regrouping, which preserves the sampled information, feasible region, optimizer, and primitive violation probability. For an equal-allocation support-subsample certificate, we derive the distortion factor \(k^2(1-ε)^{k-1}\), establish certification reversals between equivalent representations, and provide sufficient conditions for acceptance dominance under nested coarsening. We further apply nondegeneracy-free Wait-and-Judge theory to coordinatewise scenario envelopes, enabling stronger certification directly from the primitive observations. In paired scenario-MPC experiments, exact regrouping increases certification availability from 318 to 471 of 480 attempts without changing the samples, numerical proposals, or requiring any reduction in witness cardinality. Primitive Wait-and-Judge certification accepts the same 471 attempts without regrouping. These results demonstrate that apparent scenario sample-efficiency gains can arise from certificate accounting rather than additional information or reduced decision risk.

eess.SY · cs.RO · math.OC

新论文 · v1arXiv:2611.21791

Regularized Score Centering

Hao Bai, Chenlu Ye, Xuanchang Zhang, Aviral Kumar, Tong Zhang

原文摘要

Asynchronous reinforcement learning increases training throughput, but data from stale policies and numerical mismatch between training and inference can destabilize learning. To stabilize off-policy learning, score centering (SC) removes the effect of reward baselines on the expected update. Our gradient diagnostics show less extreme gradients with SC than with importance sampling (IS), but SC can still fail when the learner moves far from the behavior policy that generated the data. To limit this policy mismatch, we introduce Regularized Score Centering (RSC), which adds a reverse Kullback-Leibler (KL) penalty to encourage the learner's next-token distribution to stay close to the behavior policy's next-token distribution. RSC works with or without truncated importance sampling (TIS), which caps token importance weights. Our theoretical analysis supports this approach, showing that KL regularization improves convergence guarantees. We evaluate dense and mixture-of-experts models on mathematical reasoning and agentic tasks, including online shopping, tool use, games, and fully asynchronous software engineering. RSC with TIS substantially improves sample efficiency across all agentic benchmarks, attaining higher training rewards with the same number of sampled responses. In final evaluations, it achieves top performance across all six agentic tasks, improving by 14.3% on AppWorld and reaching 9.9 times the strongest baseline's score on BabyAI.

cs.LG

新论文 · v1arXiv:2611.21789

ASTRA: ASV Simulator for Transferable Realistic Autonomy

Tsimafei Lazouski, Cody Sheltraw, Alan Papalia, Maani Ghaffari

原文摘要

Autonomy development for marine autonomous surface vehicles (ASVs) remains challenging because field testing is costly and difficult to reproduce. Existing marine simulation tools provide mature capabilities in vessel dynamics, robotics integration, and visual rendering, but these strengths are emphasized differently across platforms. This paper presents ASTRA, an ASV simulation framework built in NVIDIA Isaac Sim that combines marine vehicle dynamics, photorealistic perception, and Robot Operating System (ROS 2) autonomy interfaces within a common environment. ASTRA incorporates a Fossen-style marine dynamics model together with geometry-based buoyancy and programmable spectral wave disturbances, enabling non-planar hydrostatic and wave-induced motion. The simulator reproduces the sensor configuration of a physical BlueBoat testbed, including GPS, IMU, LiDAR, and RGB-D sensing exposed through ROS 2. This allows localization, perception, planning, and control software to be developed against interfaces consistent with the physical platform. The BlueBoat's existing ROS 2 localization and perception modules were ported directly into ASTRA, while targeted zero-propulsion experiments were used to verify the implemented wave-body coupling. The simulated roll response was consistent with the expected non-oscillatory hydrostatic behavior, and heave remained highly coherent with the generated wave field, with approximately unity gain at the dominant wave frequency in the tested sea states. ASTRA is intended as an open-source testbed for marine autonomy development, simulation-based benchmarking, and sim-to-real experimentation.

cs.RO

版本更新 · v4arXiv:2111.11236

Nanorobot queue: Cooperative treatment of cancer based on team member communication and image processing

Xinyu Zhou

原文摘要

Although nanorobots have been used as clinical prescriptions for work such as gastroscopy, and even photoacoustic tomography technology has been proposed to control nanorobots to deliver drugs at designated delivery points in real time, and there are cases of eliminating "superbacteria" in blood through nanorobots, most technologies are immature, either with low efficiency or low accuracy, Either it can not be mass produced, so the most effective way to treat cancer diseases at this stage is through chemotherapy and radiotherapy. Patients are suffering and can not be cured. Therefore, this paper proposes an ideal model of a treatment method that can completely cure cancer, a cooperative treatment method based on nano robot queue through team member communication and computer vision image classification (target detection).

cs.RO · cs.CV · cs.LG

新论文 · v1arXiv:2611.21783

Better Calibrated, Better Planned: Uncertainty-Aware Multi-ASV Informative Path Planning for Oil Spill Monitoring

Alejandro Casado-Pérez, Marija Popović, Daniel Gutiérrez-Reina, Sergio L. Toral., Samuel Yanes

原文摘要

Monitoring oil spills with fleets of Autonomous Surface Vehicles (ASVs) requires mapping a heterogeneous pollutant field from sparse, noisy measurements under tight time and energy budgets. Informative Path Planning (IPP) addresses this problem by coupling a reconstruction model, which infers the field and its uncertainty, with a planner that decides where to sample next. Prior work has focused on planning strategies and has largely relied on Gaussian Processes (GPs), leaving the calibration of the model's uncertainty underexplored. We study how uncertainty calibration conditions multi-vehicle IPP with a fleet of ASVs that coordinates online through a shared uncertainty map. We introduce Evidential Deep Learning (EDL) as an uncertainty-calibrated information model. We benchmark EDL, Monte Carlo Dropout and Deep Ensembles against a GP in closed loop with five planners on physics-based oil spill simulations. Across 60 maps, the choice of model dominates the outcome, explaining 80\% of the variance of the final reconstruction error, against 1\% for the planner among the deep models. EDL achieves the lowest error and the highest IoU under every planner, is the best calibrated model and runs at 3.9 ms per step. With 50 cells per vehicle it reconstructs the spill better than MC Dropout and Deep Ensembles with 150. Deep Ensembles become more confident as their error grows, whereas EDL is unaffected. Robustness of the information model to the data its own planner collects is therefore a requirement of uncertainty-aware IPP.

cs.RO · cs.LG · cs.MA

新论文 · v1arXiv:2611.21782

Every Camera Grows Up: Camera-Height Regression in Interactive Video World Models

Sheng Lu, Guangpin Tao, Tong Lu, Ying Tai

原文摘要

Interactive video world models serve as simulators for embodied agents: from a starting image and an action trajectory, they generate the views the agent would see. The starting image fixes how high the camera is, for example 0.15 m for a floor robot or 2.2 m for an overhead camera, and a faithful simulator must keep this height as the camera moves. We show that six recent open-weight models fail to do so: as motion replaces the starting view with generated content, they move the camera toward human eye level, a failure we call camera-height regression. Our benchmark, EyeHeight, pairs six rendered elevations (0.15 m to 2.2 m) and four calibrated real-world rig heights with scale-free multi-view height estimation. All six models lose their initial height during a $180^\circ$ turn-and-walk rollout: initialized at 0.15 m, only 0-20% of clips end within $1.3\times$ of the starting height as inferred viewpoints converge to human heights of 1.1 m to 1.5 m, whereas ground-truth renders remain within a few percent. In Matrix-Game models, homography-tracked floor motion stays consistent with constant elevation during rotation even as newly outpainted room geometry yields a higher reconstructed viewpoint. Regression grows with rotation angle, begins while the start frame remains in context, and largely reverses in SANA-WM when a $360^\circ$ turn restores the initial view. We trace this pull to human-centric video priors in which floor-level views are nearly absent: in SANA-WM, textual height cues and dense floor-disparity maps do not prevent it, whereas classifier-free text guidance during the first half of denoising strengthens it. Combining unguided sampling with height-balanced fine-tuning recovers more retention but costs 5 MUSIQ points, a loss fixed during those same early steps. Viewpoint drift alone shifts the paths predicted by visual navigation policies.

cs.CV

新论文 · v1arXiv:2611.21781

Beyond Reconstruction: A Unified Framework for Trustworthy and Explainable EEG Artifact Removal Evaluation

Phat Lam, Hung Nguyen, Lam Pham, Huy Phan, Dat Tran, Ian McLoughlin, Roozbeh Razavi-Far, Hung Cao

原文摘要

Deep-learning methods have achieved strong performance in EEG artifact removal, but evaluation remains dominated by reconstruction metrics such as error, correlation, and signal-to-noise ratio. Although these metrics assess similarity to clean EEG, they do not establish whether models effectively suppress artifacts, preserve neural activity, or provide reliable explanations. To address these limitations, we propose TriX-EEG, a three-level framework for trustworthy and explainable EEG artifact removal. Level 1, Signal Validity (SV), evaluates reconstruction quality and neural band-power preservation. Level 2, Removal Faithfulness (RF), assesses the removed component through artifact error and residual artifact/neural leakage. Level 3, Explanation Reliability (ER), examines model behavior through temporal alignment, causal relevance, and stability. We evaluate TriX-EEG using nine published neural denoisers across three public benchmarks derived from EEGDenoiseNet and the PhysioNet Motion Artifact Contaminated EEG Dataset, covering EOG, EMG, and motion artifacts. Results show that reconstruction metrics alone are insufficient for trustworthy denoiser assessment: RF complements it by revealing whether errors stem from residual artifact or suppressed neural activity, while ER captures distinct aspects of explanation behavior, supporting joint evaluation of three aspects within a unified framework rather than reliance on any single factor. External validation on an independent EEG/EOG dataset and the BCI Competition IV-2a motor-imagery dataset further demonstrates broadly consistent rankings under dataset shift. Higher-ranked denoisers generally preserve more task-relevant information or cause less downstream performance degradation. Overall, TriX-EEG provides a comprehensive basis for EEG denoiser assessment, encouraging models that are not only accurate but also faithful and explainable.

eess.SP · cs.AI

版本更新 · v3arXiv:2605.22401

Cross-Species RSA Reveals Conserved Early Visual Alignment but Divergent Higher-Area Rankings Across Human fMRI and Macaque Electrophysiology

Nils Leutenegger

原文摘要

CORRECTION (August 2026): an evaluation-mode defect in the shared feature-extraction pipeline affected the predictive-coding and STDP conditions. It applies to both sides of every comparison here: the human values are reprinted from the companion study and the macaque values use the same checkpoints. In a single-seed re-evaluation with repaired checkpoints, macaque STDP at V1 moves from 0.305 to 0.266 and PC from 0.210 to 0.178, level with the untrained baseline, so finding (2) holds for STDP but not for PC; the V4 and IT orderings change, while V2 is unchanged. Revision note: the human V1 results are not robust to the analysis choices examined in arXiv:2611.15533. Findings (1), (3) and (4) are unaffected, and other conditions change by at most 0.0034. Five-seed analysis, tau, noise ceilings and stimulus control not re-computed. See the correction note on page 1; the abstract below is unchanged. Follow-up study: arXiv:2608.12408. Does the relationship between learning rules and brain alignment generalize across species? We test the same five learning rules against macaque electrophysiology. The macaque data come from MajajHong2015 (V4/IT, 3,200 presentations, 88/168 neurons) and FreemanZiemba2013 (V1/V2, 135 stimuli, 102/103 neurons). Using RSA with identical model weights from our human study, we find: (1) all models achieve higher alignment with macaque early visual cortex (rho = 0.15-0.30 at V1/V2) than with human fMRI (rho = 0.01-0.08); (2) STDP and PC produce the highest macaque V1/V2 alignment; (3) at IT, rankings show no detectable correlation across species (Kendall's tau = 0.00), though this null is expected given that n = 5 provides power only at tau = +/-1.0; (4) a pretrained ResNet-50 achieves rho = 0.25 at macaque IT, substantially above all custom CNN conditions (rho = 0.07-0.14), suggesting IT alignment is limited by model capacity rather than by the learning rule.

cs.LG · cs.NE · q-bio.NC

新论文 · v1arXiv:2611.21777

LLMs as Validators for Cross-Chain Bridges: Evaluating Source-Level and Transaction-Level Attack Detection on Solana

Yasir Rahman, Ishraq Samiha, Shahnewaz Karim Sakib, Anindya Bijoy Das

原文摘要

Cross-chain bridges have suffered substantial asset losses due to vulnerabilities in their on-chain programs. On Solana, missing validation checks can enable exploits even when the underlying blockchain operates correctly. The ability of large language models (LLMs) to diagnose these flaws from source code or transaction records, rather than merely raise alarms, remains insufficiently understood. We implement a native Rust lock-and-mint bridge on a local Solana validator with four exploitable variants: two model flaws from the Cashio and Wormhole incidents, while the other two represent vulnerability classes documented in Solana security audits. Paired vulnerable and secure programs, together with successful attacks, benign sessions, and blocked attempts, allow us to distinguish correct diagnoses from unsupported alarms. We evaluate six configurations of five open-weight models across 16,800 responses, compare source inputs with and without comments, and grade 7,200 answers on vulnerable inputs using two blinded LLM graders. The results reveal substantial differences across conditions: without comments, models diagnose 24% of vulnerable source inputs while frequently flagging the secure bridge. Transaction diagnosis is high for the missing-signer scenario but averages 6% across the other three scenarios. Original comments, some of which describe the flaw, increase source diagnosis to 60%. These findings highlight the importance of distinguishing correct vulnerability diagnoses from unsupported alarms in bridge security review.

cs.CR · cs.AI

版本更新 · v4arXiv:2609.30789

Feature Selection for Tabular Foundation Models: Exploring Feature Utility in Low-Data Molecular Prediction

Yiqi Yao, Miquel Duran-Frigola

原文摘要

Choosing features is difficult with few labeled molecules, even for a tabular foundation model that learns in context. We study this problem with TabPFN for low-data molecular property prediction, separating input width, feature meaning, and predictive usefulness while keeping contexts, query molecules, and predictor settings matched. On ADME/Tox assays from Therapeutics Data Commons, endpoint-matched assay predictions improve a compact RDKit descriptor baseline, and a provenance audit shows that most audited query molecules already occur in the predictors' training sources. Across 77 FS-Mol development tasks and 157 held-out test assays drawn from ChEMBL, broad RDKit descriptor and fingerprint families provide a strong default, rank-averaging TabPFN with support-fitted fingerprint forests improves upon both models, and learned weighting of family predictions adds further gains at a larger prediction budget. A utility-trained selector builds a traceable 16-column input from 238 sourceassay predictors, improving upon the baseline while trailing wider representations and equalwidth structural SVD coordinates. Named outputs identify what each feature predicts and where it comes from. Their usefulness for a new assay is assessed alongside the other input features, and paired removal and replacement tests measure what each selected output contributes. Together, these studies offer a protocol for evaluating feature selection, representation choice, and data provenance with a fixed tabular foundation model for few-shot molecular property prediction.

cs.LG

版本更新 · v2arXiv:2610.26836

Same Simulation, Different Evidence: Where Retrieved Tool Output Costs LLM Agents Accuracy

Ke Zhang, Sahchit Chundur, Jerry Peprah Owusu, Mohammad Javad Abdolhosseini Qomi, Maziar Raissi

原文摘要

LLM agents that compute the right answer can still answer wrongly. Agent harnesses increasingly keep long tool outputs out of the context window, saving them to files that the agent searches or reads on demand. We ask where this costs accuracy with a real scientific tool: PHREEQC, a widely used geochemistry simulator from the USGS, whose reports reach 846.8k characters and whose determinism lets every printed value be traced. Agents use it to answer 158 execution-verified multiple-choice items. When agents read the simulator's report through a section index rather than in context, GPT-5.1 and GPT-5.2 lose 11.3 and 7.9 points of accuracy, while Claude Opus 4.6 and Claude Sonnet 4.6 show no detectable cost. Re-running all 11,007 logged simulator calls shows that both GPT models compute the keyed value about as often as before; what changes is the evidence they answer from. Under retrieval they often skip the report, answer from their own higher-precision output files, and misround to a neighboring option, a failure we call evidence substitution. With the simulation held fixed, retrieval still costs GPT-5.2 9.4 points. For GPT-5.4, more frequent correct simulations mask an evidence-use loss of about ten points, leaving a net change indistinguishable from zero. Naming the authoritative output raises accuracy by 6.3-7.9 points over plain retrieval for the three GPT models; for GPT-5.1 and GPT-5.2, inlining outputs that fit works as well, while asking for correct rounding does not. Context-management choices in agent harnesses should be judged by where accuracy is lost, not only by final accuracy.

cs.AI

新论文 · v1arXiv:2611.21772

Boxes in Infinite Words: Complexity, Automata, and Point Modules

Jason Bell

原文摘要

Motivated by the geometry of point modules over monomial algebras, we introduce a box-counting function associated to a right-infinite word over a finite alphabet. An extendable box is a family of finite words obtained by choosing letters independently at each position, subject to the condition that all choices admit a common infinite right continuation in the associated subshift. We prove several results for these functions for words of linearly bounded complexity. For $k$-automatic words, we show using techniques from first-order logic that the total box-counting function is $k$-regular. Finally, we apply our results to prove eventual affine growth of the number of positive-dimensional irreducible components in finite projections of point-module spaces associated to semiprime monomial algebras of at most quadratic growth.

cs.FL · math.LO · math.RA

新论文 · v1arXiv:2611.21769

Learning Calibrated and Samplable Predictive Distributions of Surface-Ocean $Δp$CO$_2$ Using Tabular Diffusion Models

Gabriele Accarino, Mohammad M. Amirian, Thea H. Heimdal, Amanda R. Fay, Galen A. McKinley, Viviana Acquaviva

原文摘要

The ocean absorbs almost 30% of anthropogenic CO$_2$ emissions, but quantifying this sink requires estimating surface-ocean CO$_2$ from observations that are sparse in space and time. Most existing machine-learning estimates of this field are deterministic, giving a single value per grid point, with uncertainty approximated post hoc via residuals, ensembles, or prescribed distributions. We instead learn a conditional predictive distribution for each observation: using Treeffuser, a conditional diffusion model with gradient-boosted-tree score functions, we estimate the full predictive distribution of $Δp$CO$_2$ (the sea-air $p$CO$_2$ difference) from 13 environmental, spatial, and temporal predictors. On a held-out SOCAT test set, Treeffuser attains the lowest continuous ranked probability score among the uncertainty-quantification methods evaluated (8.229 $μ$atm) and the coverage closest to the nominal 68% / 95% levels (0.686 / 0.930), while retaining point-prediction skill comparable to a deterministic XGBoost baseline (RMSE 18.4 vs. 17.7 $μ$atm; $R^2$ 0.77 vs. 0.79). These results show that conditional diffusion can provide well-calibrated, observation-level uncertainties while largely preserving point-prediction skill. Because Treeffuser is generative, its samples can support probabilistic analyses of $Δp$CO$_2$; extending this to spatially coherent field reconstruction will require modeling dependence across grid cells and evaluation on spatially held-out data.

cs.LG · physics.ao-ph

版本更新 · v3arXiv:2605.30556

Supervised Training Rapidly Degrades Early Visual Cortex Alignment Across Biologically Plausible Learning Rules

Nils Leutenegger

原文摘要

CORRECTION (August 2026): the central finding of this paper is not supported. An evaluation-mode defect left the batch-normalisation layers of the predictive-coding and STDP conditions in training mode during feature extraction, producing their apparent preservation of V1 alignment. With the defect repaired, predictive coding degrades V1 alignment more than backpropagation does, not less. The finding that training degrades V1 alignment for every rule tested does survive. [Revision note: not robust to the analysis choices examined in arXiv:2611.15533.] See the correction note on page 1; the original abstract below and the body are unchanged from v1. Corrected analysis: arXiv:2608.12408. Revision note: see Comments. Random, untrained neural networks consistently match or exceed trained networks in representational similarity to early visual cortex. This puzzling finding challenges the assumption that learning improves brain alignment. We investigate it by tracking representational similarity analysis (RSA) alignment to human fMRI data across training for four learning rules: backpropagation (BP), feedback alignment (FA), predictive coding (PC), and spike-timing-dependent plasticity (STDP). Using 720 object images from the THINGS database and fMRI data from three subjects across six visual ROIs, we measure Spearman correlations between model and brain representational dissimilarity matrices at eight training checkpoints (epochs 0-40). We find that (1) a single epoch of training reduces V1 alignment by 25-90%, depending on the learning rule; (2) backpropagation reduces V1 alignment most severely (delta r = -0.080), while predictive coding and STDP preserve substantially more (delta r ~ -0.04); and (3) a weaker, opposite tendency appears in object-selective cortex (LOC), where BP shows the largest increase in alignment during training, although the absolute change is small.

cs.LG · q-bio.NC

版本更新 · v5arXiv:2604.16875

Untrained CNNs Exceed Backpropagation in V1 Alignment at High Evaluation Resolution: A Systematic RSA Comparison of Four Learning Rules Against Human fMRI

Nils Leutenegger

原文摘要

Revision note: the V1 result reported here is not robust to the analysis choices examined in arXiv:2611.15533; see page 1. We compare four learning rules (backpropagation (BP), feedback alignment (FA), predictive coding (PC), and spike-timing-dependent plasticity (STDP)) in identical convolutional networks against human fMRI from THINGS-fMRI (720 stimuli, 3 subjects) using representational similarity analysis (RSA), together with an untrained random-weights baseline that isolates the contribution of architecture. Models are trained on 32 px CIFAR-10 and evaluated at 224 px; results are averaged across 5 seeds, with uncertainty from resampling stimuli. At V1, the untrained baseline exceeds BP rho = 0.053$ vs. $0.022$; rho = +0.031$, 95% CI $[0.019, 0.043]$), and the same holds at V2. This advantage depends on the evaluation resolution: in the 5-seed model set of the resolution study (gap $+0.030$ at 224 px), it vanishes when the models are evaluated at their 32 px training resolution rho = -0.001, 95% CI $[-0.011, 0.009]$; arXiv:2608.12408). At LOC, BP, PC and STDP exceed the untrained baseline, FA is borderline, and the trained rules do not differ from each other. At IT, no two conditions can be distinguished at the low reliability of the data (leave-one-subject-out lower bound 0.021). FA has the lowest alignment at V1. At V1, an untrained network thus matches BP at the training resolution and exceeds it at 224 px; learning-rule differences are resolvable only at an intermediate area.

cs.LG · q-bio.NC

版本更新 · v2arXiv:2608.11056

Twisted Conjugacy and the Construction of Induced Centrosymmetric Alternant Codes

Ousmane Ndiaye, Massamba Sow

原文摘要

This paper presents a construction of induced centrosymmetric alternant codes based on the automorphism structures inherited from Generalized Reed-Solomon (GRS) codes. We introduce a twisted conjugacy relation naturally arising from projective semilinear transformations and establish its connection with ordinary conjugacy in the corresponding projective semilinear coset. By means of the Shintani norm, we show that, in the quadratic extension case, all admissible transformations inducing centrosymmetry have a trivial Shintani norm and therefore belong to a unique conjugacy class in the semilinear coset. We further characterize the explicit forms of these transformations and establish a transport theorem describing how twisted conjugation acts on the supports and multipliers of GRS codes. These results provide an algebraic framework for constructing and analyzing induced centrosymmetric alternant codes.

cs.IT · math.CO

版本更新 · v4arXiv:2608.12408

Evaluation Resolution Confounds Learning-Rule Comparisons in Model-Brain RSA of Early Visual Cortex

Nils Leutenegger

原文摘要

Revision note: see Comments. Studies of biologically plausible learning rules (feedback alignment, predictive coding, STDP) train small networks on $32\times32$ images and compare them with brain responses to naturalistic stimuli presented at much higher resolution. We find that a common result in this setting, that untrained or locally trained networks rival or beat backpropagation at early visual cortex, depends on the resolution at which networks are evaluated. In representational similarity analysis (RSA) against human V1, the gap between an untrained and a backprop-trained network grows monotonically from $-0.001$ (95% CI $[-0.011, 0.009]$) at the 32 px training resolution to $+0.030$ $[0.018, 0.042]$ at 224 px (six resolutions, 5 seeds; per-subject RSA on cross-run stimulus pairs). The growth reproduces for three further learning rules, directionally in single-seed macaque electrophysiology, along training, and for an ImageNet ResNet-50 and a Swin-Tiny transformer, which also align best at low resolution although trained at 224 px. Gabor and pixel structure, the normalization state of the untrained baseline, and a global brightness statistic do not account for it. Against a criterion fixed before the analysis, limiting image content to 32 px while the input still grows does not shrink the V1 gap ($+0.035$ $[0.023, 0.046]$ vs. $+0.030$ at 224 px; ratio 1.14 $[1.03, 1.34]$): it arises once the input exceeds the training resolution, through a decline of backprop. A single luminance value per image reaches rho= 0.054 against V1, matching the untrained network (0.053), which bounds what RSA on globally pooled features can resolve at V1 in this dataset. The one learning effect that holds across resolution is backprop above untrained at LOC. Comparisons of learning rules or architectures at early visual cortex need to control, and report, the evaluation resolution.

q-bio.NC · cs.LG

新论文 · v1arXiv:2611.21766

Margin Dynamics for Large Language Model Alignment

Xingzi Xu, Mehmet Saygin Seyfioglu, Karim Bouyarmane

原文摘要

Practitioners choose among LLM alignment methods largely by trial and error. We provide a theoretical foundation: each pairwise method defines a vector field on the probability simplex that governs how the model redistributes mass between preferred and dispreferred completions. Projecting onto a single preferred/dispreferred pair reduces this to a one-dimensional margin ODE with a closed-form solution. These solutions reveal that DPO margins grow without bound (logarithmically as $O(\log t)$ in the gradient-flow limit) while IPO margins converge to a finite $1/(2τ)$; that label noise induces a qualitative phase transition in DPO (unbounded $\to$ bounded) but affects IPO only quantitatively; and that gradient concentration under asymmetric weighting governs trainability at scale. We confirm all three predictions across five model families spanning four benchmarks. The closed-form solutions also enable principled method design: we prove that adding any label smoothing $\varepsilon > 0$ to SimPO creates a finite margin attractor at $Δ^* = β^{-1}\log((1-\varepsilon)/\varepsilon)$, while $\varepsilon = 0$ (standard SimPO) has no attractor. We test this across three models (8B-9B) and confirm the prediction: the theory-derived configuration (TPO-S, $\varepsilon = 0.05$) is competitive with the strongest baselines, attaining the highest mean IFEval score among the compared methods on all three, validating that the predicted finite attractor is a usable operating point rather than proposing a new state-of-the-art loss.

cs.LG · cs.AI

新论文 · v1arXiv:2611.21751

From Top to Bottom: Quark Mass Hierarchies in Heterotic String Theory

Yuri Chervonyi, Fabian Ruehle, Nenad Tomasev

原文摘要

Standard Model fermion masses exhibit large hierarchies, spanning more than five orders of magnitude in the quark sector alone. In this paper, we study whether these hierarchies can be generated perturbatively at tree level in smooth heterotic Calabi-Yau compactifications, where the supergravity approximation remains under computational control. In such backgrounds, physical Yukawa couplings combine holomorphic superpotential integrals with matter field Kahler metrics. The former can be computed algebraically, while the latter require solving a coupled system of partial differential equations for the Ricci-flat metric, Hermitian Yang-Mills connections, and harmonic bundle-valued forms. Within this framework, we identify two new tree-level mechanisms by which complex structure moduli can generate fermion mass hierarchies and show analytically and numerically that both mechanisms are obstructed for previously studied tetra-quadric models. Motivated by these obstructions, we scan the existing databases of tens of thousands of complete intersection Calabi-Yau line bundle models and find a unique smooth quotient admitting a rank-three up-quark Yukawa matrix at tree level. Evaluating the physical couplings on this manifold and evolving them under the renormalization group, we obtain realistic mass hierarchies for the top, charm, up, and bottom quarks in the smooth supergravity regime.

hep-th · cs.AI

新论文 · v1arXiv:2611.23329

A level is not a rank: an empirical audit of quantile-to-index conversion in 10 released conformal-prediction libraries

Shivam Lalakiya

原文摘要

Context. A conformal interval at 0.90 keeps that guarantee only at the right rank: a fresh score falls at or below the rth smallest of n with probability r/(n + 1). The libraries we study instead hand a level to a quantile function, which can land elsewhere. At n = 20, the upper end of statsforecast's interval sits at position 18.05 where rank 19 is required. Objective. Where does installed code resolve the wrong rank, at what cost in coverage, and what does it return when no rank reaches the level? Method. We audit 39 sites where a level becomes an index in 10 pinned packages, a named set rather than a systematic search. 12 are classified by running them, 11 through their shared numpy method, the rest from source. Coverage is paired against the required rank on the same scores. Results. All 7 tabular implementations reach the required rank. In darts, on one code path, coverage is unchanged (0.0000) where the ranks agree and 0.0280 to 0.0560 lower where they differ; a second archive repeats that split but not its range. Below the feasibility floor, $n < 1/α- 1$, of 14 public call paths 5 silently return a finite number; the rest raise or warn. Conclusions. Quantile interfaces take only a level; statsforecast's default calibrates below its floor. 6 of 19 upstream filings have merged, one fixing a coverage shortfall. A conformance suite classifies helpers by the rank they reach, and a checklist guides review.

cs.SE

新论文 · v1arXiv:2611.21608

潜在格语言模型

Latent Lattice Language Models

Shivam Duggal, Zongze Wu, William T. Freeman, Antonio Torralba, Richard Zhang, Eli Shechtman

AI 速读 基于摘要

作者提出在Transformer每层增加沿时间循环的横向潜在流(Lattice of Latents),以增加活跃计算状态而非深度或循环层数,在20B FineWeb-Edu token上降低验证困惑度。

cs.LG · cs.AI · cs.CL

查看原文摘要

Language models typically increase per-token computation through deeper or looped layers and latent reasoning steps, while richer residual connections improve information flow across layers. We instead increase computational capacity: the number of active computational states a model maintains. We introduce the Lattice of Latents, a set of per-layer horizontal latent streams that run recurrently along time, alongside the vertical residual stream that flows through depth. Like hyper-connections, mHC, and attention or value residuals, the Lattice enhances the residual stream, but rather than adding or rerouting residual pathways through depth, it provides companion computational states that persist across tokens. Unlike hybrid Transformer-SSM architectures, whose recurrent states primarily serve as internal memory for sequence mixing, these latents directly condition the vertical stream's attention, KV-cache writes, MLP computation, and decoding. On Transformers trained on 20B FineWeb-Edu tokens, a 406M-parameter Lattice model lowers validation perplexity from 15.41 to 14.40 over a parameter-matched Transformer. These gains extend across architectural variants. The Lattice composes with prior residual-stream enhancements: with hyper-connections and input residuals, it lowers validation perplexity from 15.41 (base Transformer) and 14.22 (Transformer with the same enhancements) to 13.61. On Gated DeltaNet, Mamba-2, Triadic GDN, and GDN-Mamba-2 hybrids, it improves perplexity at every tested native state size, with Lattice-GDN and Lattice-Triadic-GDN variants outperforming pure baselines with 4x larger native states. These results identify persistent per-layer state as a complementary axis for scaling language-model computation.

新论文 · v1arXiv:2611.21607

ScaleVLA:将视觉-语言-动作模型扩展到5520亿参数

ScaleVLA: Scaling Vision-Language-Action Models to 552 Billion Parameters

Ruixi Yu, Dairu Liu, Xuchuan Chen, Zekun Qi, Zekai Li, Zhuo Cao, Luyang Niu, Yingsheng Zhang 等 12 位作者

AI 速读 基于摘要

作者将552B的DeepSeek-V4.1-Flash视觉语言骨干接入可训练的动作DiT,并在统一训练设置下比较2B到552B的VLM对机器人策略的影响。

cs.RO

查看原文摘要

Frontier Vision-Language Models (VLMs) such as GPT-6 Astra have demonstrated remarkable capabilities in robot control, suggesting that broad pretraining provides knowledge valuable for physical action. Yet most Vision-Language-Action (VLA) policies still rely on much smaller VLM backbones, leaving this potential largely untapped. We ask whether frontier-scale representations can provide a stronger foundation for robot learning. We introduce ScaleVLA, which connects a 552B DeepSeek-V4.1-Flash backbone to a trainable action DiT, and systematically compare VLMs from 2B to 552B under a shared training setup. We evaluate three fundamental capabilities of robot foundation models. In sample efficiency, just 10 demonstrations outperform a 2B backbone trained with 50 (87.8% vs. 85.6%). In generalization, LIBERO-PRO success reaches 45.1%, compared with at most 31.3% for smaller backbones. In precise manipulation, real-world charger insertion improves from 4/10 to 9/10 successes, with gains concentrated in contact-sensitive alignment. Despite its scale, ScaleVLA achieves 216 ms median latency per action chunk. Our findings suggest that frontier VLMs can serve not only as reasoning agents, but also as powerful representational foundations for efficient and generalizable robot policies.

版本更新 · v3arXiv:2605.18476

AI4BayesCode: From Natural Language Descriptions to Validated Modular Stateful Bayesian Samplers

Jungang Zou, Alex Ziyu Jiang, Qixuan Chen

原文摘要

Coding and computation remain major bottlenecks in Markov chain Monte Carlo (MCMC) workflows, especially as modern sampling algorithms have become increasingly complex and existing probabilistic programming systems remain limited in model support, extensibility, and composability. We introduce AI4BayesCode, an extensible LLM-driven system that translates natural-language Bayesian model descriptions into runnable, validated MCMC samplers. To improve reliability, AI4BayesCode adopts a modular design that decomposes models into modular sampling blocks and maps each block to a built-in sampling component, reducing the need to implement complex sampling algorithms from scratch. Reliability is further improved through pre-generation validation of model specifications and post-generation validation of generated sampler code. AI4BayesCode also introduces a novel recursively stateful coding paradigm for MCMC, allowing modular sampling components, potentially developed by different contributors, to be composed coherently within larger MCMC procedures. We develop a benchmark suite to evaluate AI4BayesCode for sampler-generation. Experiments show that AI4BayesCode can implement a wide range of Bayesian models from natural-language descriptions alone. As an open-ended system, its capability can continue to expand with improvements in the underlying AI agent and the addition of new built-in blocks. The software is available at https://ai4bayescode.com/.

stat.CO · cs.AI · cs.LG

新论文 · v1arXiv:2611.21604

FaceJev: Direct Visual Decisions for Face Forgery Detection

Siran Peng, Tianshuo Zhang, Yibo Xue, Ajian Liu, Haoyuan Zhang, Weisong Zhao, Shuhuan Chen, Xiangyu Zhu 等 9 位作者

原文摘要

Face forgery detection methods often perform well on seen manipulations but generalize poorly to unseen ones. Multimodal Large Language Models (MLLMs) can recognize facial parts and assess their visual plausibility. Existing MLLM-based methods typically express these judgments through textual explanations, but seldom treat individual part decisions as explicit learning targets. We present FaceJev, an MLLM-based method for face forgery detection through direct visual decisions. Following the Jev formulation, FaceJev answers eight independent questions about facial parts using a shared image encoding and reads anomaly probabilities from the native answer logits. A face is classified as forged if any part is predicted to be abnormal, directly linking the detection result to the part decisions. We construct PartForensics, a dataset of approximately 160k images with binary anomaly targets for the eight facial parts, to supervise these decisions. We further propose FS-GRPO, which optimizes each part decision through an exact reward comparison with its alternative, using both detection and part rewards. FaceJev outperforms the strongest baseline by 7.25 percentage points on unseen manipulations. Extensive ablation studies validate our design choices and the effectiveness of the main components.

cs.CV

新论文 · v1arXiv:2611.21605

ViRA: Toward Ultra-High-Resolution Visual Agents via Cross-Input Relative Policy Optimization

Huawei Lin, Jinting Liu, Tongyan Wang, Xuyang Chen, Ruihua Shi, Zhichao Xu, Jiaxin Qin, Guoshenghui Zhao 等 12 位作者

原文摘要

Visual agents answer questions about ultra-high-resolution photos by zooming into the relevant regions. However, existing agents often answer from common sense instead of reading the evidence. Reinforcement learning (RL) on the final answer cannot correct this, because a guess earns the same reward as reading. We propose Cross-Input Relative Policy Optimization (CrossRPO), which trains agents on contrast sets. A contrast set asks one question about several versions of an image that differ only in a small region with the evidence and have different answers, so only reading the evidence solves the whole set. CrossRPO rewards solving the whole set, and we prove that it does not reinforce a set answered from common sense. We build contrast sets for training (ViRA-Train) and evaluation (ViRA-Bench) by editing one small object in images of 21 to 120 megapixels, and we photograph further sets with real changes (ViRA-Bench-Real) to exclude editing traces. We train two Visual Reasoning Agents, ViRA-9B and ViRA-27B, with CrossRPO on only about 3,500 items of ViRA-Train each. CrossRPO achieves 63.3% set accuracy on ViRA-Bench at the maximum resolution, against 59.3% for GRPO. ViRA-9B is the most accurate open model of its size on all benchmarks, and ViRA-27B outperforms Claude Haiku 5.5 and GPT-6-Luna, both at their highest thinking level, on most benchmarks.

cs.CV · cs.AI

新论文 · v1arXiv:2611.21602

量化人类预报员讨论对热带气旋强度预测的附加价值

Quantifying the added value of human forecaster discussions for tropical cyclone intensity prediction

Ilia Azizi, Frederick Iat-Hin Tam, Marc-Olivier Boldi, Valérie Chavez-Demoulin, Tom Beucler

AI 速读 基于摘要

该研究分析2002-2023年637个热带气旋的讨论文本,发现其与SHIPS预测因子结合可降低24小时最大持续风速预报的平均绝对误差。

cs.LG · cs.CL

查看原文摘要

The National Hurricane Center routinely releases Tropical Cyclone Discussions that offer insights into its tropical cyclone forecasts. This study examines the added value of these discussions for intensity forecasting beyond a set of numerical predictors. For 637 Atlantic and eastern North Pacific tropical cyclones from the 2002-2023 hurricane seasons, we analyze the discussion text and its combination with 13 hurricane environmental predictors from SHIPS and ERA5 10-meter wind fields, for forecasting intensity changes at lead times of 24 to 120 hours using three modeling methods. In one approach, an end-to-end deep learning framework for the text shows that the discussions alone offer comparable value for 24-hour intensity forecasts to the SHIPS predictors. Adding discussion information to the SHIPS predictors in an ensemble of five such models reduces the mean absolute error of 24-hour maximum sustained wind forecasts for 2022-2023 cyclones from 9.89 to 8.90 kt, with a confidence interval that excludes zero when whole storms are resampled. The ERA5 wind fields have a minor effect on mean absolute error in wind predictions, and model rankings differ across predicted variables, lead times and model types. We provide prediction intervals for these forecasts, whose coverage falls below the 90% target during rapid intensification. Analysis of the text shows that words related to intensity and intensity change (such as intensify, weaken, eyewall and category) are important for the model predictions. This retrospective study shows the added value of the forecasters' discussions and suggests that multi-modal learning can help understand the predictive contribution of expert human judgment. The reanalysis data differ from real-time analyses, and the discussions are issued at least three hours after the valid time of the other inputs, so these results should not be interpreted operationally.

新论文 · v1arXiv:2611.21599

Hyperlyve: Parallelized Enumeration and Exact Lyapunov Stability Certification for Deep ReLU Networks

Brian Yarbrough, Jesse Wayment, Jingbo Wang, Philip E. Paré, Shreyas Sundaram

原文摘要

We introduce Hyperlyve (Hyperplane Partitioned Lyapunov Verification and Enumeration), a framework for proving the asymptotic stability of nonlinear dynamical systems, including those with neural network controllers. Hyperlyve learns and formally certifies continuous piecewise-affine (CPA) neural Lyapunov functions within a counterexample-guided inductive synthesis (CEGIS) loop. Hyperlyve's key innovation is a parallel algorithm that enumerates the hyperplane arrangement formed by the ReLU activations of deep networks and certifies that Lyapunov conditions hold within each polyhedral region. We derive sound and complete regional tests to verify the asymptotic stability of the exact smooth dynamics (rather than a piecewise-linear approximation) within a compact domain of interest. Furthermore, we formally characterize the computational complexity of the CPA neural Lyapunov verification problem as co-NP-complete, filling a key gap in the literature. Empirical results demonstrate that Hyperlyve enumerates ReLU networks over 150-times faster than competing tools, which contributes to Hyperlyve proving the asymptotic stability of systems that are currently intractable for state-of-the-art verifiers.

eess.SY · cs.CG · cs.LG · cs.LO · math.DS

新论文 · v1arXiv:2611.21596

SADL: On the Demand-Supply Geometry of Representation Collapse

Junhan Sun, Yigui Feng, Mingzhe Liu, Haoyan Xu, Lei Ye, Ziqi Han, Hanru Bai, Yanzhao Zhang 等 10 位作者

原文摘要

In Joint-Embedding Predictive Architectures (JEPAs), a well-constrained projector output does not by itself prevent representation collapse. We find that, from initialization, the encoder representation is dominated by a component shared by different inputs. A batch-normalized projector removes this component, so regularization after the projector cannot correct it, and the differences between representations shrink until they collapse. We propose SADL (Supply And Demand Loss), which keeps the original training objective and constrains the encoder directly when needed. SADL compares the representation differences that the projector output demands with the differences the encoder can retain and pass on to the projector, and adds a constraint on the encoder only when supply falls short of demand. It works with different Demand Regularizers and introduces no auxiliary networks, additional trainable parameters, or architectural changes. In single-task and jointly trained four-task LeWorldModel (LeWM) and INTACT experiments, SADL lowers the collapse rate over ten random seeds from 50% to 0%. After one epoch of training, encoder effective rank is 3.4-6.4 times that of the corresponding baseline without SADL, and latent prediction error falls to 1/20-1/303 of it. Without additional tuning, SADL stays robust as its loss weight spans orders of magnitude, and with several Demand Regularizers INTACT exceeds 95% mean search-free control success. On four image datasets, SADL also consistently improves accuracy and stability across random seeds. These results show that matching the representation differences a regularizer demands with the differences the encoder can supply leads to faster and more stable representation learning in control and image classification.

cs.RO · cs.LG

新论文 · v1arXiv:2611.21590

The Provenance-Privacy Paradox: Can Private Cloud Compute (PCC) Prove That Apple Reference Images Are Not Used for Learning?

Mohammad I. Husain, Viet Nguyen

原文摘要

Apple Reference Image (ARI) combines camera-sensor signatures, Secure Enclave attestation (SEP), cryptographic timestamp bounds, Private Cloud Compute (PCC), and a final post-quantum signature to establish photographic provenance. PCC is central to this process because it validates the secure digital negative and develops authenticated raw sensor measurements into a reference JPEG. Although the secure digital negative is encrypted during transmission, the authorized PCC workload must process the raw pixels in plaintext. This requirement creates a provenance-privacy paradox: cloud processing strengthens the provenance guarantee while giving an Apple-authorized, machine-learning-capable workload access to valuable, hardware-authenticated raw imagery. We argue that workload attestation and transparency-log inclusion verify software identity and authorization but do not, by themselves, prove that an authorized workload cannot learn from or retain information derived from the photograph. We implement this counterexample in a Python- and Docker-based experimental testbed that models relevant ARI and PCC functions. Across 27 public-domain photographs, a baseline workload and a learning-enabled workload both passed the same verification checks and returned byte-for-byte identical JPEGs. However, the learning-enabled workload retained a first-layer gradient of approximately 48 KB from which its 64 x 64 input was reconstructed with a maximum absolute per-pixel error of 6 x 10^-8. The additional processing overhead was small relative to image decoding and processing. These results do not demonstrate that Apple currently trains on ARI photographs or reproduce production PCC, but identify non-learning and non-retention as independently verifiable assurance requirements.

cs.CR