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

+0新增论文
+0其中 AI 相关
+0已有论文修订
+92新增 AI 导读

arXiv 最近完成扫描:2026-11-23 01:07 · 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 2026-11-22 01:35 — 2026-11-23 01:35(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:2026-11-23 01:07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 35 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-23 01:05(北京时间)
新论文 · v1arXiv:2611.26159

When Foundation Models Disagree: Representational Heterogeneity for Calibrated Uncertainty in Whole-Slide Image Classification

Yekai Shen, Yuqian Wang, Zishun Liao, Rongsheng Wang, Wenting Chen, S. Kevin Zhou, Qingsong Yao

原文摘要

Pathology foundation models supply strong representations for whole-slide image classification, yet the confidence attached to their predictions remains hard to trust. Prevailing uncertainty estimators act within a single feature space and thereby overlook the uncertainty that the choice of foundation-model representation itself induces. This work introduces CALIFUSE, a calibration-oriented framework that fuses several pathology foundation models for the specific purpose of uncertainty estimation. Departing from accuracy-driven fusion, CALIFUSE reads representational heterogeneity across encoders as an explicit confidence cue. Slide-level decision consensus is coupled with chunk-level evidence agreement quantified through Jensen-Shannon divergence, so that both prediction stability and fine-grained evidence consistency inform the estimate. Foundation-model encoders stay frozen. A task-specific nnMIL classifier is still trained for each FM; CALIFUSE then estimates confidence at inference without training an extra fusion module. Evaluated on seven WSI classification tasks under several foundation-model combinations, the framework sharpens calibration, misclassification detection, and selective prediction relative to both the strongest single-encoder baseline and naive ensemble averaging, all while preserving competitive balanced accuracy.

cs.CV · q-bio.QM

新论文 · v1arXiv:2611.25643

TRAC: Efficient Diffusion Language Models for Machine Translation via Length-Aware Decoding

Tianyuan Li, Xihan Yang, Chenglong Wang, Yuzhang Wu, Jingbo Zhu, Tong Xiao

原文摘要

Machine translation with large language models requires both high translation quality and low generation latency. Accelerating translation while preserving quality is therefore essential for practical deployment. Diffusion large language models (dLLMs) can generate tokens at arbitrary target positions through iterative denoising, offering a new paradigm for parallel translation beyond sequential autoregressive decoding. However, existing methods still incur redundant computation: fixed target generation lengths waste computation on unused positions, while confidence-only commitment delays medium-confidence tokens whose predictions have largely stabilized. To address these issues, we propose TRAC, a training-free adaptive decoding method for diffusion machine translation. TRAC adapts the target generation length using source length and translation direction, with a conservative fallback to the maximum length for empty or unusually short outputs. It preserves high-confidence commitment while identifying temporally stable medium-confidence tokens for early commitment using cross-step top-1 persistence and approximate Jensen-Shannon divergence over Top-$K$ distributions. On bidirectional Chinese-English translation across FLORES-200, WMT23, and NTREX-128, TRAC reduces model forward calls by 24.9% to 36.8% and achieves sentence-level generation speedups of 1.91 to 2.19 times over Fast-dLLM, with only a 0.02-point decrease in macro-average COMET. These results demonstrate that adaptive generation lengths and stability-aware token commitment improve the efficiency of diffusion machine translation.

cs.CL

新论文 · v1arXiv:2611.25168

What a Calibrated Probability Hides: Belief-Level Post-Training for Decision Models

Xinye Li, Zirui Song, Huaxing Liu, Xiang Wang, Shuai Li, Zixiang Xu, Yinjun Liu, Zhewen Zhang 等 10 位作者

原文摘要

A calibrated decision model can predict how often each answer is chosen. The same prediction can reflect disagreement among people who are confident in their choices or uncertainty among people who share a belief. Disagreement calls for identifying whose perspective applies, while uncertainty calls for gathering more evidence. Standard post-training scores a single prediction against one observed choice at a time, learning the pooled distribution without distinguishing these two sources of uncertainty. We introduce Belief-Grouped Policy Optimization (BGPO). The policy proposes a list of beliefs, estimates their prevalence, and predicts a choice distribution for each. During training, BGPO uses Bayes' rule to estimate how likely each belief was to generate an observed choice. These estimates weight the choice's contribution to each belief's log-score reward. To discover rare beliefs, the policy also proposes lists after seeing a few observed choices and retains them only if they predict held-out choices well. At inference, the policy uses only the question and combines the belief-specific distributions by prevalence to produce a calibrated aggregate prediction. On benchmarks covering opinions, conversation safety, emotions elicited by artwork, and photo aesthetics, BGPO achieves the lowest conditional, worst-group, and overall Jensen-Shannon divergence across model initializations. It improves conditional divergence on both text and image tasks. On a synthetic image task with known beliefs, BGPO recovers the underlying groups more accurately than baselines trained with belief labels.

cs.CV

新论文 · v1arXiv:2611.24511

OmniFish: A Bioinspired, Modular, Reconfigurable Soft Robotic Fish Enabling Multi-Modal Swimming

Mathias Jensen, Saravana Prashanth Murali Babu

原文摘要

In this work, we introduce OmniFish, a bioinspired modular soft robotic fish designed to emulate the full spectrum of body and/or caudal fin (BCF) locomotion modes within a single robotic platform. The design combines morphology averaging across representative fish species with a tendon-driven compliant segmented spine whose effective stiffness can be reconfigured through progressive mechanical jamming. By selectively constraining body segments, OmniFish shifts deformation from whole-body undulation to posterior body bending and localized tail oscillation, enabling bioinspired Anguilliform, Subcarangiform, Carangiform, Thunniform, and Ostraciiform swimming using a single tendon-driven actuator. The design was guided by morphometric analysis and evaluated through caudal-fin thrust characterization, numerical body-drag simulations, kinematic analysis, and swimming performance experiments. Kinematic analysis confirmed the intended robot transition from distributed body undulation to localized tail oscillation. Swimming experiments demonstrated distinct mode-dependent performance, with the Ostraciiform configuration achieving the highest straight-line velocity of 8.14 cm/s and the lowest cost of transport of 2.42 at 1.5 tail beats/s. These proof-of-concept results show that a mechanically simple stiffness-reconfigurable architecture can generate distinct BCF swimming modes and provide a versatile strategy for developing multimodal soft robotic fish.

cs.RO · physics.flu-dyn

新论文 · v1arXiv:2611.23417

Autoregressive View Model

Jensen, Zhou, Samuel Rota Bulo, Denis Rozumny, Duncan Zauss, Peter Hedman, Nikhil Keetha, Philip Torr 等 11 位作者

原文摘要

We present the Autoregressive View Model (AVM), which generates a scene as a stream of views without constructing an explicit 3D representation. Through a unified in-context interface, AVM conditions on images or videos to generate short clips along a requested camera path. Visual context grounds reconstruction and interpolation, while a learned prior enables extrapolation into unseen regions. AVM retrieves relevant views from inputs and generated history to form a bounded active context. We fine-tune a pretrained bidirectional video diffusion transformer with flow matching in an N-in-1-out formulation. We then tune a causal model on its own rollouts to improve photorealism, reduce denoising steps, and limit long-horizon drift. We scale training to approximately 8M scenes and video clips. We evaluate generation capacity and scene persistence across benchmarks for camera control, text control, visual quality, dynamics, sparse and 4D reconstruction, and long-horizon consistency. The results demonstrate that AVM adapts to diverse input and output settings through one in-context interface.

cs.CV

新论文 · v1arXiv:2611.20602

Quantum Computing-Based Wasserstein Generative Adversarial Networks for Airfoil Parameterization

Abdur Rahman Ovy, Xiaosong Du

原文摘要

Airfoil geometry parameterization is essential in aerodynamic design optimization. Optimization search in the large space suffers from low efficiency and modest feasible region, which could further incur failures to convergence. Generative adversarial networks (GAN), aiming to learn the statistical distribution of existing data, pave a new avenue for shape generation and airfoil parameterization. GAN manages to implicitly reduce design space by generating only smooth, realistic airfoils. Meanwhile, quantum computing has been emergingly exhibiting great potential for achieving superior performance in specific types of engineering problems but has not been introduced to airfoil parameterization. To fill this gap, this work develops quantum computing-based Wasserstein GAN (Q-WGAN) and investigates the performance for airfoil parameterization. The Q-WGAN adopts parameterized quantum circuits as the generator to promote the generative performance, while the discriminator remains classical. The developed Q-WGAN is demonstrated on a dataset of optimal airfoils parameterized by free-form deformation (FFD) control points, resulting from 992 airfoil aerodynamic design optimizations. Results show that Q-WGAN achieves 92.1% mean fitting accuracy on the optimal FFD control points and 99.2% accuracy on airfoil coordinates, comparable with the classical WGAN performance of 92.28% on FFD control points and 99.26% on airfoil coordinates. Q-WGAN achieves lower Jensen-Shannon divergence, Fréchet distance, and Wasserstein distance than the classical WGAN, indicating an improved agreement with the reference airfoil distribution. Therefore, this work presents a convincing evidence that Q-WGAN makes a viable option for parameterizing airfoils and possesses a great potential for other relevant engineering applications.

cs.LG

新论文 · v1arXiv:2611.18511

近乎相等分布间散度的数值问题

On numerical issues of divergences between nearly equal distributions

Brandon Amos

AI 速读 基于摘要

作者指出分类分布间常用散度(前向/反向KL、广义JSD、蒙特卡洛KL估计与熵)在分布近乎相等时存在舍入与相消的数值问题,并提出数值稳定的重写形式。

cs.LG · math.NA · stat.ML

查看原文摘要

Barron et al. (2026) recently identified numerical issues in the softmax and soft cross-entropy loss. We show that similar issues are present in standard divergence operations between categorical distributions. This includes the forward and reverse Kullback-Leibler (KL) divergence and generalized Jensen-Shannon divergence (JSD), as well as Monte Carlo KL estimates and entropy. When the distributions are nearly equal, rounding and cancellation issues affect the forward and gradient computations. These issues can affect settings using divergences in optimization objectives, such as knowledge distillation, reinforcement learning, and proximal methods. We propose numerically stable reformulations and provide example implementations for each setting, validated by property-based tests.

新论文 · v1arXiv:2611.16270

SMIC: Shared Multiscale Input-Reference Conditioning across Transformer Layers for Natural Language Inference

Xin Zhang, Victor S. Sheng

原文摘要

Natural Language Inference models can be sensitive to localized lexical changes. We introduce Shared Multiscale Input-Reference Conditioning (SMIC), which constructs a single input-dependent multiscale reference and reuses it across Transformer layers. Each layer applies its own multiscale transformation to the shared reference and injects the resulting conditioning feature before the attention update. We instantiate SMIC in a token--context Transformer (TCT) and conduct from-scratch comparisons against decoder-based and Multi-Layer Perceptron (MLP) baselines across all three ANLI rounds. TCT-SMIC attains the highest observed ten-seed mean accuracy on all three ANLI rounds. The ablations further support the contribution of SMIC. We evaluate a real-valued Llama-style plug-in and extend deletion/insertion diagnostics to pretrained ModernBERT encoder and Llama decoder models with and without SMIC. Across the evaluated perturbation protocols, TCT-SMIC exhibits comparatively low pooled-attention redistribution and smaller signed accuracy changes. We further establish a finite-sample bound on average subset-mass redistribution in terms of the mean Jensen--Shannon divergence between the restricted and renormalized attention distributions.

cs.CL · cs.AI · cs.LG

版本更新 · v2arXiv:2611.14284

Ninety-Six Ways to Measure a Scribe: Fragmented Evaluation of Ambient Scribes in Healthcare

Amama Mahmood, Shiye Cao, Laura Fleig, Junan Jiang, Theodore Jensen, Yee-Yin Choong, Kristen Greene, Chien-Ming Huang

原文摘要

Ambient artificial intelligence (AI) scribes have moved rapidly from experimental pilots to enterprise-wide deployment across health systems, promising reduced documentation burden and improved clinician experience. While research evaluating these tools has expanded quickly, there are no agreed-upon protocols or standardized metrics for assessing their performance, making it difficult to compare findings or cumulatively synthesize evidence across studies. Across 54 recent evaluation studies, we identify ten evaluation dimensions and 40 distinct metrics, operationalized in 96 different ways-35 of them (36%) used by only a single study. Assessment is dominated by documentation efficiency and time, measured in 44 of 54 studies. The other dimensions appear at broadly similar rates (16-24 studies each), and safety and clinical outcome least consistently of all. Ripple effects on the broader care system, such as impact on patients' care outcomes, redistribution of work, changes in billing and coding intensity, and downstream use of notes by other clinicians are rarely assessed. Moreover, the metrics used rely heavily on self-report and are fragmented and inconsistently operationalized, which limits comparability and hinders synthesis across studies. As an effort to support more rigorous and cumulative evaluation, in this paper we review the study designs, data sources, and metrics commonly used to evaluate ambient AI scribes in healthcare settings, discuss the limitations of current evaluation practice, and argue for a more holistic approach grounded in a standardized, multi-dimensional set of metrics.

cs.HC · cs.AI

版本更新 · v2arXiv:2611.13498

Evaluating Compositional Generalization in Transformers: The Role of Composition Equivalence and Module Coverage

Purva Pruthi, Andrew Yuan, Alexander D'Amour, David Jensen

原文摘要

Compositional generalization is an important form of out-of-distribution generalization, defined as the ability to train on some combinations of modules (e.g., functions such as sort, filter) and then generalize to unseen combinations. Despite substantial empirical work evaluating compositional generalization in transformers, findings have been inconsistent because evaluations were often conducted on a limited set of train-test splits, defined in an ad hoc manner. Through a systematic evaluation, we identify two key aspects of the train-test distribution that strongly affect performance: composition equivalence and module coverage. We show that the apparent performance of direct models (trained only on final outputs) results primarily from exploiting composition equivalences, meaning different sequences of modules that reduce to identical end-to-end functions. When benchmarks eliminate these equivalences, performance drops significantly, showing their inability to generalize to compositions of known modules that produce novel end-to-end functions. We further demonstrate that performance degrades monotonically as module coverage shifts increase (when modules appear in different contexts between train and test), affecting both direct and step-by-step models (trained on intermediate and final outputs). These findings have concrete implications for benchmark design: benchmarks that include compositional equivalences can lead to significant overestimation of the generalization abilities of direct models, and benchmarks that include both module coverage violations and compositional equivalences can exacerbate shortcut learning. As a case study, we demonstrate the role of compositional equivalence on the generalization performance of the pre-trained model Llama3-8B on the code-output prediction task in the coding benchmark LiveCodeBench.

cs.LG · cs.CL

版本更新 · v2arXiv:2603.01750

Deep Heteroskedastic Regression: Post-Hoc Variance Estimation from Latent Representations

Mikkel Jordahn, Jonas Vestergaard Jensen, James Harrison, Michael Riis Andersen, Mikkel N. Schmidt

原文摘要

Uncertainty quantification (UQ) in deep learning regression is of wide interest, as it supports critical applications including sequential decision making and risk-sensitive tasks. In heteroskedastic regression, where the uncertainty of the target depends on the input, a common approach is to train a neural network that parametrises the mean and the variance of the predictive distribution. Yet to this day, training deep heteroskedastic regression models poses severe practical challenges in the trade-off between uncertainty quantification and mean prediction, such as variance overfitting, optimization difficulties and representation collapse. In this work, we identify and discuss the core issues and propose a simple and efficient procedure that addresses these jointly by post-hoc fitting a variance model across the intermediate layers of a pretrained network on a hold-out dataset. We demonstrate that this method is competitive with end-to-end trained mean-variance networks in heteroskedastic UQ on several data modalities. The method retains mean prediction accuracy, is cheap at both train and prediction time, and requires no additional data compared to existing methods. Finally, we show this method works on large scale foundation models for chemistry, paving the way for cheap heteroskedastic UQ in large scale neural networks without having to retrain such models end-to-end.

cs.LG

版本更新 · v2arXiv:2602.05026

Laws of Learning Dynamics and the Core of Learners

Inkee Jung, Siu Cheong Lau

原文摘要

We formulate an algorithm-independent theory of learning dynamics for individual models and ensembles. We establish a conservation law, the vanishing of predictive entropy under successful learning, quantitative entropy and core-expansion bounds, and calibration and finite-sample validation results for identifying reliable predictions. These results motivate a two-generation ensemble that uses Jensen-Shannon (JS) disagreement to route uncertain inputs to specialized learners. CIFAR-10 experiments examine controlled perturbations, distribution shifts, and adversarial attacks. On a heterogeneous test environment, routing improves accuracy over probability averaging while keeping core accuracy near 95%. For adversarial defense, we consider a query-only black-box setting: the attacker can query the overall prediction function but does not know a particular internal decomposition into learners and routing rules. Our current attacks use a fixed-initial-route gradient, which is accurate for a single-step attack when the route remains unchanged, but may deviate from actual multi-step attacks, especially for large perturbations. We evaluate our two-layer defense against this approximation; whether additional adversarially trained layers improve resistance to stronger multi-step attacks remains for future study.

cs.LG · cs.AI · cs.IT

版本更新 · v7arXiv:2502.12756

Navigating Demand Uncertainty in Container Shipping: Deep Reinforcement Learning for Enabling Adaptive and Feasible Master Stowage Planning

Jaike van Twiller, Yossiri Adulyasak, Erick Delage, Djordje Grbic, Rune Møller Jensen

原文摘要

Reinforcement learning (RL) has successfully solved various deterministic and stochastic planning problems. However, conventional RL struggles with complex real-world constraints, particularly when feasibility is explicit and depends on the current state or trajectory. In this work, we address stochastic sequential decision-making with state-dependent constraints through a real-world case study of the master stowage planning problem in container shipping, which aims to optimize revenue and costs under demand uncertainty and operational constraints. We propose a deep RL framework with an encoder-decoder model that integrates problem instance, solution, and uncertainty information to guide planning. We introduce differentiable projection layers that enforce convex polyhedral constraints, while Jacobian corrections offset the projections to yield unbiased policy gradient estimates. Experiments show that our model efficiently finds adaptive, feasible solutions that generalize across distribution shifts and scale to longer planning horizons, outperforming state-of-the-art baselines in constrained RL and stochastic programming. As such, our policies enable adaptive, uncertainty-aware planning that can support resilient and sustainable supply chains.

cs.LG · math.OC

新论文 · v1arXiv:2611.10641

Same Contacts, Different Rooms:Meeting Format Shapes LLM Group Influence Beyond Pairwise Contact Counts

Zhexi Gu, Shusheng Cao, Runping Zhu

原文摘要

Simulations of LLM agent collectives usually summarise interaction as a pairwise contact network. We test that summary with a contact-matched intervention: the members, their initial stances and every pairwise contact count stay fixed while the meetings are reorganised, and each member gives a one-step support readout after reading the arguments of everyone present. Across six open-weight LLMs and 14 debate topics, merging triads into five-person meetings (resize) moves group support more than reshuffling who meets in triads (regroup), although both move the same 60 dyadic exposures. The root-mean-square change is 4.36 versus 2.59 percentage points (pp), larger for resize in all six models and in 81 of 84 model-topic cells. Resizing also has a consistent direction: it amplifies the initial majority of the merged members by 1.72 pp, in every model, whereas regrouping has no direction (<= 0.03 pp). A contact network cannot represent regrouping at all, yet pairwise rules applied meeting by meeting explain up to 76% of its effect, and with identical weights they raise resize R^2 by 0.51-0.58 for two models. We show that the gap between the two placements is an exact Jensen gap of the sigmoid readout, bounded by how much a member's predicted logits vary across meetings; this dispersion tracks the gap with log-log r=0.97 and tells a modeller when contact counts are a safe summary. Meeting format is a first-class variable for LLM social simulation.

physics.soc-ph · cs.CL

版本更新 · v2arXiv:2605.07020

FlashMol: High-Quality Molecule Generation in as Few as Four Steps

Xinyuan Wei, Zian Li, Shaoheng Yan, Cai Zhou, Muhan Zhang

原文摘要

Generating chemically valid 3D molecular conformations is critical for computational drug discovery. Classical diffusion-based models like GeoLDM perform well but require hundreds of steps, making large-scale in silico screening impractical. Recent efforts on few-step molecular generation have accelerated this process to 12-50 steps, but they often largely sacrifice sample stability. In this work, we present FlashMol, an ultra-fast molecule generative model producing high-quality molecular conformations in as few as 4 steps. To achieve this, we adapt distribution matching distillation (DMD) -- a reverse KL-divergence minimization objective -- to the molecular domain for effective distillation. Considering the local minimization behavior of DMD, we respace the molecule generation timesteps, providing the generator with much better initialization and enables effective distillation. Additionally, to mitigate the mode-seeking behavior of DMD and improve diversity, we further regularize it with a Jensen-Shannon divergence term, which incorporates the mean-seeking behavior of the forward KL divergence. Extensive experiments on QM9 and GEOM-DRUG datasets demonstrate that FlashMol matches and even surpasses the original 1000-step teacher: by applying respacing to the GeoLDM teacher (GeoLDM-respaced), we achieve comparable results to the 1000-NFE teacher with as few as 16 NFEs. After further distillation, FlashMol is able to sample with only 4 NFEs while maintaining high molecular quality, achieving up to 250$\times$ acceleration compared to the GeoLDM teacher.

cs.LG · cs.AI

版本更新 · v2arXiv:2609.05522

Diffusion models for eye-gaze trajectory generation using position and velocity representations

Laxman Basnet, Alexander Szorkovszky, Pedro G. Lind, Anis Yazidi, Shailendra Bhandari

原文摘要

Eye-tracking data are expensive to collect, requiring specialized hardware and controlled laboratory conditions, and difficult to share due to privacy regulations, which together limit the quantity and diversity of gaze data available for research. We investigate unconditional gaze generation using diffusion models in position and component-wise velocity representations. The models are trained and evaluated on eight-second visual-search trajectories from 28 participants, using a split of 23 training participants and five held-out validation participants, with three training seeds. Evaluation includes kinematic distributions, trajectory geometry, event statistics, and a temporal Wasserstein GAN with gradient penalty as a learned comparator. Across eight position-derived features, mean Jensen-Shannon divergence is $0.016 \pm 0.004$ for the position model and $0.037 \pm 0.002$ for the velocity model, with standard deviations calculated across training seeds. These scores do not establish that representation choice caused the difference. A separate comparison using the same evaluation method shows that the ordering of the models varies across training seeds. The position model also produces smaller feature discrepancies than a kinematic Markov baseline. Differences in saccade counts between real and generated data depend strongly on the velocity threshold used to detect events. The results provide quantitative reference measurements for unconditional gaze generation on this visual-search task and show how evaluation choices affect conclusions about generated gaze.

cs.CV · cs.AI · cs.NE

版本更新 · v3arXiv:2609.22547

When Pass@$k$ Crossovers Are Real - and Why: Paired Inference and the Markov Kernel of RLVR

Chen Yang

原文摘要

Reinforcement learning with verifiable rewards (RLVR) typically improves pass@1 but can fall behind its base model at large sampling budgets $k$, commonly interpreted as sharpening existing solutions. Because adjacent pass@$k$ estimates reuse generations, pointwise error bars cannot establish a crossover. We formulate paired inference on the RL-minus-base difference curve: a prompt-level multiplier bootstrap yields a one-sided dominance test and simultaneous confidence bands requiring both an early gain and a later loss. Among nine public RLVR checkpoints at 3k-16k token limits, only one narrowly meets this criterion before multiplicity adjustment; asymmetric truncation can create, mask, or reverse crossovers. On 1,060 deduplicated competition problems at 32k tokens, two 1.5B models exhibit certified crossovers around $k\approx10$-$60$, while two 7B models show no supported loss. A paired variance decomposition shows why adding prompts improves power more than deeper sampling. Prompts with identical base success probability diverge after RLVR, with dispersion replicating across independent generation halves and violating Lipschitz slope bounds for reasonable deterministic maps. Modeling RLVR as a Markov kernel, Jensen's inequality shows that conditional dispersion depresses pass@$k$ for $k\ge2$. Once dispersion is modeled, the conditional-mean log-odds slope is below one in every pair: RLVR improves hard prompts on average, while late-$k$ loss comes from a 2%-3% minority of base-solved prompts collapsing onto a repeated modal wrong answer and resisting prefix rescue up to 4,096 tokens. In controlled runs from one base model on 3,000 new problems, all four GRPO arms fall behind near $k\approx200$ and remain behind through $k=4096$, while SFT stays significantly ahead throughout.

stat.ML · cs.LG

版本更新 · v2arXiv:2606.16988

Agent trajectories as programs: fingerprinting and programming coding-agent behavior

Hamidah Oderinwale

原文摘要

Benchmark scores tell you what an agent got right; they do not tell you how it got there. In this work, we introduce methods for comparing agents procedurally in different contexts, where the model, tasks, and approaches vary. We compare ten agents and find that they are identifiable by their behavioral habits, which we define as fingerprints: a probe over these procedural signatures attributes an unseen trajectory to the correct agent at 85.7% accuracy. We develop procedural representations for agent problem-solving procedures with an emergent vocabulary induction technique which is meant to be maximally compressive to avoid surface-level variation while being expressive enough to unveil the quirks of the models' patterns. We apply our framework to the software engineering evaluation dataset SWE-Bench to study the structural distinctness of agent trajectories and find that behavior is most similar between models from similar release periods and those that are distilled from one another (e.g., a distilled student model and its teacher have a Jensen-Shannon divergence of 0.25, about half the distance between other model pairs). As more models saturate evaluations, we believe that it will be important to probe model behavior along more holistic dimensions than success rates alone. We introduce ProcGrep, a library for auditing and evaluating agents for how they approach tasks at a procedural level given their traces in a top-down fashion. ProcGrep is available at https://github.com/hamidahoderinwale/procgrep.

cs.SE · cs.LG

新论文 · v1arXiv:2611.01651

Physics-Informed One-Step Generation on Long-short flow maps

Zhiqi Li, Duowen Chen, Jinjin He, Bo Zhu

原文摘要

Physics-informed generative models should produce samples that are both data-aligned and physically valid, yet existing methods are often iterative and too slow for practical use. Achieving this with a one-step flow-map model is difficult because both data-distribution and physics-constraint learning are enforced only indirectly through surrogate objectives, making their joint learning fragile for long-range flow maps. We propose \textbf{Physics-Informed Long-Short Flow Maps with Direct Guidance (LSFM-DG)}, which makes direct target computation feasible by decomposing one-step generation into a long flow map for global evolution and a near-terminal short flow map with computable closed-form solutions and physics potentials. As the short flow map shrinks to zero, it provides both direct distributional and physical targets for learning the long flow map. This yields a unified one-step flow-map model that removes Jensen's gap, improves training stability, and achieves stronger data fidelity and physical consistency without distillation.

cs.LG

新论文 · v1arXiv:2611.01378

Information-Theoretic Design of Sparse Autoencoder-Based Steering

Hirofumi Morise

原文摘要

Sparse autoencoder (SAE)-based activation steering is a lightweight approach for modifying the behavior of large language models (LLMs) by adding small vectors to hidden states during inference. However, designing effective steering interventions remains challenging. In this work, we develop an information-theoretic framework for analyzing SAE activations through empirical feature-firing distributions. The central object of the framework is the Jensen--Shannon divergence (JSD) between such firing distributions, which quantifies how far a generated or intervened state deviates from a dataset-induced target distribution. Based on an analytic divergence response for virtually shifted states, we construct steering vectors toward the target firing distribution. The same divergence can also be used to monitor deviation from another empirical distribution, such as a general valid-language distribution. We demonstrate the effectiveness of the steering in a question-mode generation task study. We further use teacher-forced pulse experiments to show that local interventions induce structured downstream JSD responses in later layers, and apply the framework to a safety-related target--contrast response setting for feature analysis.

cs.CL · cs.LG

新论文 · v1arXiv:2610.40566

全维高斯混合模型 NPMLE 的自正则化与结构自适应

Self-regularization and structural adaptation of NPMLE for Gaussian mixture models in all dimensions

Yanjun Han, Yandi Shen, Yihong Wu

AI 速读 基于摘要

作者证明:均值位于超立方体内的 d 维高斯混合模型,其 NPMLE 能以高概率自动拟合出 O((log n/log log n)^d) 个分量,并在有限混合数据下使拟合分量数依概率有界。

math.ST · stat.ML

查看原文摘要

We establish structural properties of the nonparametric maximum likelihood estimator (NPMLE) for $d$-dimensional Gaussian mixtures whose means lie in a hypercube. We show that, with high probability, the NPMLE based on $n$ independent observations self-regularizes by fitting a mixture model with $O((\frac{\log n}{\log\log n})^d)$ components, achieving the optimal model complexity without explicit model selection. Furthermore, the NPMLE exhibits a structural adaptation property: For data drawn from a finite mixture, the number of fitted components is bounded in probability as $n\to\infty$. Our proof applies a multivariate Jensen-type bound to control the typical number of critical points of the KKT function associated with the first-order optimality conditions. The key step is to lower bound its gradient norm uniformly on complex spheres. We establish this lower bound by decomposing the KKT function into an empirical fluctuation term and a relative density estimation error term, combining small-ball probability estimates for the former with polynomial approximation and Hellinger error bounds for the latter.

新论文 · v1arXiv:2610.38100

Evaluating Causal Reasoning in LLMs: A Case Study in the Challenges of Procedural Generation

Justin Clarke, Isha Karn, Patrick Murray, KJ Hill, David Jensen

原文摘要

Large language models (LLMs) appear to represent and reason about a large range of phenomena. In particular, recent studies have examined whether LLMs can accurately reason about causation. One of the most commonly used benchmarks for evaluating formal causal reasoning in LLMs is CLadder, a framework for procedural generation of benchmark data sets that was released in 2023. Experiments with CLadder appear to indicate that LLMs can perform formal causal reasoning in a large percentage of generated examples. In this paper, we examine what can be reliably inferred when LLMs perform well on test sets produced by CLadder. We identify several examples of superficial and highly predictive patterns that the CLadder generator embeds in data sets it generates. We find that even simple models such as random forests perform well on CLadder-generated data sets after training on a sample of generated data. We also find that the performance of many commercial LLMs on CLadder test sets is extremely brittle and that removing large sections of the text that are critical for formal causal reasoning has almost no effect on performance. Given these results, the most likely hypothesis for the strong performance of state-of-the-art LLMs on CLadder is that they are exploiting superficial patterns learned from exposure to training data from the generator. Any inference that LLMs are exhibiting more complex causal reasoning must rule out this hypothesis.

cs.AI · cs.LG

新论文 · v1arXiv:2610.37868

Human-Residual Steering: Aligning Frozen LLM Judges with Human Judgment Distributions

Chen Lyu, Phong Do, Gabriele Pergola

原文摘要

Large language models (LLMs) are widely used as automated judges, yet most alignment methods target aggregated human judgments and overlook variation among individual annotators. Existing approaches to matching the full distribution of human ratings mainly rely on output-level calibration or parameter fine-tuning. We propose Human-Residual Steering (HRS), to our knowledge the first method to align frozen LLM judges with human judgment distributions through learned hidden-state interventions. An oracle analysis shows that low-rank corrections at a single layer remove most of the discrepancy between judge and human distributions. HRS learns to predict such corrections from the judge's hidden representation and rating logits, training on individual annotations with proper scoring rules for ordinal and nominal judgments and optimising through the frozen downstream layers. Across 14 benchmarks and four judges from three model families, HRS outperforms all evaluated frozen-weight baselines. With Qwen2.5-7B, it achieves the lowest Jensen-Shannon distance on 13 of 14 benchmarks, reducing the mean from 0.565 to 0.352 versus 0.377 for the strongest calibrator. HRS also raises mean Spearman correlation with aggregate human scores from 0.449 to 0.579, improves population-level estimation under limited annotation budgets, and trains about 88 times fewer parameters than distributional fine-tuning at a modest cost in alignment.

cs.CL

新论文 · v1arXiv:2610.36984

Quantum Evolution Kernel for Node-Attributed Graphs

Manuel Padovani, Alessandra Di Pierro, Claudia Daffara

原文摘要

We present a novel approach to embed node-attributed graphs onto neutral-atom quantum computers, and propose an application to digit recognition. Our approach extends PASQAL's quantum evolution kernel (QEK) by introducing a node-level feature encoding based on three-level quantum states. To this purpose, we present a prototype showing how the resulting embedding can be leveraged to extract correlation-based information for classification tasks. The entire quantum process is simulated using Pasqal's Pulser library, which inevitably imposes constraints on the size of the experimental datasets. We compare our method with classical baselines to evaluate its potential. We show that, for the digit recognition task, our method achieves a higher mean accuracy than the more commonly used Jensen-Shannon distance-based methods by utilizing the Wasserstein distance.

quant-ph · cs.LG

版本更新 · v2arXiv:2610.31850

Rethinking Multi-View Fusion for Urban Representation Learning

Sean Bin Yang, Zongyi Xu, Jilin Hu, Ying Sun, Kristian Torp, Bin Yang, Torben Bach Pedersen, Christian S. Jensen

原文摘要

Urban region representation learning encodes heterogeneous urban signals, such as mobility flows, points of interest, and land-use patterns, into unified embeddings for downstream tasks, including check-in prediction, crime forecasting, and service demand estimation. Existing approaches rely on globally shared fusion and joint optimization, implicitly assuming uniform modality importance and naturally balanced multi-objective training. However, urban data is inherently heterogeneous. Our empirical analysis reveals two structural issues: (i) modality utility varies markedly across regions, causing static fusion to overemphasize signals from certain modalities; and (ii) heterogeneous objectives exhibit persistent gradient imbalance, causing representational bias and unstable training. We present PRISM, a self-adjusting framework for adaptive multi-view learning. PRISM adapts sparse Mixture-of-Experts (MoE) routing to learn region-specific view allocations, while a Proportional-Integral-Derivative (PID)-inspired regulation mechanism mitigates gradient conflicts during optimization. Experiments on three large-scale urban datasets show that PRISM achieves more competitive performance than strong multi-view baselines, while providing more stable optimization behavior and more balanced region representations.

cs.LG · cs.AI

新论文 · v1arXiv:2610.35982

Generation and Verification: Human-Agent Collaboration for Safety-Critical Automotive Software Engineering

Elizaveta Zinovyeva, Aiham Taleb, Nikita Kozodoi, Daniel Schleicher, Amir Mahdi Namazi, Martin Kraus, Levent Kent, Brian Jensen

原文摘要

Research on coding agents mainly focuses on generation quality. Yet, safety-critical automotive software makes two additional demands. It prioritizes verification, since every change is reviewed, traced to a requirement, and checked against industry standards such as MISRA and AUTOSAR before it can be committed. It is also a narrow domain, written against layered architectures, embedded resource constraints, and organization-specific interface conventions that are scarce in the data general-purpose LLMs are trained on. We present a coding assistant for automotive C/C++ that addresses both demands. First, it exposes three tools to the user over the Model Context Protocol (generate, edit, and verify), allowing developers to run verify on any code on demand rather than relying on the agent to check its own output. Second, rather than using a general-purpose model, we customize the underlying LLM for the automotive domain through prompt engineering and fine-tuning. We construct a dataset from automotive production code of AUMOVIO, where each function is paired with the requirement texts and Unified Modeling Language (UML) behavior it must satisfy. In a two-week pilot on active production projects, developers report verify as the most-used tool and confirm using it on their own code and not only on generated output. Comparing model customization strategies, 75% prefer the faster prompt-engineered configuration. MISRA- and AUTOSAR-inspired pattern checkers indicate that prompt engineering achieves a lower number of violations on C, while fine-tuning shows better conformance with architecture patterns for C++.

cs.SE · cs.AI

新论文 · v1arXiv:2610.35032

AutoPRL: Automated Path Representation Learning via a Learnable View Generator

Sean Bin Yang, Ying Sun, Yunyao CHeng, Hao Miao, Jilin Hu, Bin Yang, Torben Bach Pedersen, Christian S. Jensen

原文摘要

Self-supervised learning has emerged as a powerful paradigm for Path Representation Learning (PRL) in road networks. PRL aims to learn informative representations of paths traversed by moving objects, such as vehicles, where each path consists of a sequence of connected road segments. These representations are essential for various downstream tasks, including travel-time estimation and path ranking. However, existing methods often suffer from ineffective view generation and inefficient training procedures due to their reliance on heuristic augmentations and computationally intensive contrastive learning. To address these issues, we propose AutoPRL, an effective and efficient self-supervised PRL framework that integrates three key components. The Automatic Data Augmentation (AutoDA) module employs variational autoencoders to generate semantically meaningful and diverse feature-level path views by modeling each topology-aware node embedding with a Gaussian latent distribution, thereby mitigating information loss and enhancing representation quality. Second, the Cross-View Decoupled Contrastive (CVDC) learning mechanism improves training efficiency by disentangling the optimization of positive and negative pairs across node- and path-level views, enabling more stable contrastive learning. Third, an Adaptive Co-Training (ACT) module dynamically adjusts uncertainty-based loss weights to balance multiple objectives, ensuring efficient and scalable training. Extensive experiments on three real-world datasets demonstrate that AutoPRL achieves strong and consistent performance in both effectiveness and efficiency across multiple downstream tasks.

cs.LG · cs.AI

新论文 · v1arXiv:2610.34689

RSKD: Grounding Semantic Consistency in Driving VLMs via Distillation

Chuchu Xie, Yang Zhong, Kangan Qian, Hao Ye, Kun Jiang, Diange Yang

原文摘要

Knowledge distillation can transfer large multimodal models to compact models for autonomous driving, but uniform token-level matching can dilute planning-relevant semantics and amplify teacher errors. We propose RSKD, a semantic-consistency-guided distillation framework that routes supervision according to both the task importance of an answer phrase and its semantic consistency with the teacher prediction. Task-critical semantic phrases are localized in the ground-truth answer, and a phrase-level consistency score controls the balance between Jensen--Shannon divergence (JSD) distillation and ground-truth supervision. Teacher information is emphasized only where it is semantically consistent, while semantically inconsistent phrases receive stronger ground-truth supervision. We evaluate RSKD on DrivingVQA, LingoQA, and MapLMv2. RSKD improves over the compared student-based baselines on DrivingVQA and LingoQA and remains competitive on MapLMv2.

cs.CV

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-demo.41ACL Anthology · ACL 2026

TokCollate: A Comprehensive Tool for Tokenizer Evaluation and Visualization across Languages

Dušan Variš Charles University Prague, Abishek Stephen, Jindřich Libovický Charles University Prague

原文摘要

Tokenization quality varies significantly across languages, contributing to disparities in LLM performance and cost for speakers of less-resourced languages – a phenomenon known as the “token premium” problem. Despite growing research interest, no existing tool provides a comprehensive intrinsic evaluation of tokenizers paired with interactive visualization. We present TokCollate (pronounced similarly to chocolate), a Python-based evaluation framework combined with a JavaScript visualization interface that addresses this gap. TokCollate implements a wide range of intrinsic metrics, including monolingual measures such as average token length and Rényi/Shannon efficiency, and cross-lingual measures such as vocabulary overlap, Jensen-Shannon divergence, alignment-based Eflomal scores, and length ratios. It further enables analysis across language groups defined by genealogical families, scripts, geographic regions, speaker populations, and estimated data availability. TokCollate is open-source under the MIT license and available on GitHub.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1334ACL Anthology · ACL 2026

N-GLARE: An Non-Generative Latent Representation-Efficient LLM Safety Evaluator

Zheyu Lin, Jirui Yang, Yukui Qiu, Yubing Bao, Hengqi Guo, Yao Guan

原文摘要

Evaluating the safety robustness of LLMs is critical for their deployment. However, mainstream Red Teaming methods rely on online generation and black-box output analysis. These approaches are not only costly but also suffer from feedback latency, making them unsuitable for agile diagnostics after training a new model.To address this, we propose N-GLARE (A Non-Generative, Latent Representation-Efficient LLM Safety Evaluator). N-GLARE operates entirely on the model’s latent representations, bypassing the need for full text generation. It characterizes hidden layer dynamics by analyzing the APT (Angular-Probabilistic Trajectory) of latent representations and introducing the JSS (Jensen-Shannon Separability) metric.Experiments on over 40 models and 20 red teaming strategies demonstrate that the JSS metric exhibits high consistency with Red Teaming safety rankings at less than 1% token and runtime cost.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表