跳到正文
研踪PaperTrail

版本更新

查看已发布论文的最新修订,在详情页对照实际观测到的版本。

部分采集异常

全站更新概况

+1084新增论文
+883其中 AI 相关
+130已有论文修订
+94新增 AI 导读

arXiv 最近完成扫描:2026-11-18 04:28 · 全站最近入库:2026-11-18 19:45各信源进度 →

统计口径与数据新鲜度

统计 2026-11-18 05:43 — 2026-11-19 05:43(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:2026-11-18 04:28最近发现新增版本:2026-11-18 19:45库内论文最新更新时间:2026-11-18 08:00

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

26,870 已收录论文22,374 AI 相关8,156 有修订112 HF 社区发现

论文修订列表

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

找到 8,156 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-18 04:27(北京时间)
版本更新 · v2arXiv:2608.06112

从孤岛算法到合规优先的智能体平台:医院AI系统的多层架构

From Siloed Algorithms to Compliance-First Agentic Platforms: A Multi-Layered Architecture for Hospital AI Systems

Manideep Dhar, Ritwik Singh, Sharat Chandra Kumar Manikonda

AI 速读 基于摘要

提出一种面向医院、合规优先的多层智能体AI架构,并用合成数据与原型演示分诊、流程优化与合规日志的端到端编排。

cs.AI · cs.CL · cs.LG · cs.MA

查看原文摘要

Hospitals are rapidly adopting artificial intelligence for triage, imaging, scheduling etc., yet most deployments remain isolated point solutions locked inside departmental silos, resulting in duplicated effort, hidden risks, and unrealized enterprise value. Despite explosive growth of AI in healthcare market and accelerating investment, an estimated 70-80% of healthcare AI pilots fail to scale, largely due to governance gaps, fragmented data, and missing integration blueprints. This research proposes a hospital-specific, compliance-first, Agentic AI architecture with multiple interoperable layers, extending existing hospital AI platform models with: (i) an Agent Orchestration Layer for multi-agent workflows across clinical, operational, and financial domains, (ii) a Compliance and Policy Layer that centralizes policy-as-code for HIPAA, GDPR, the EU AI Act, DISHA Act, India's DPDP Act, and ISO/IEC security and safety standards, and (iii) a Privacy-Preserving Data Fabric that plugs federated learning, differential privacy, and secure enclaves into real-world Hospital Information Management System (HIMS) flows. Using a synthetic but structurally realistic hospital dataset and an open, ready-to-deploy prototype implementation, this study demonstrates the end-to-end orchestration of triage risk prediction, workflow optimization, and compliance logging, achieving substantial simulated reductions in task turnaround times and manual documentation effort while maintaining policy-guarded data access. The resulting architecture offers hospital leaders a pragmatic blueprint to move from ad hoc tools to a governed, globally compliant, ROI-focused AI platform that can be tailored to on-premise, hybrid and cloud-native deployments.

版本更新 · v2arXiv:2610.01376

重构 React 组件层级以消除 Prop Drilling

Refactoring React Component Hierarchies to Eliminate Prop Drilling

Vangelis Gkinis, Vassilis E. Zafeiris

AI 速读 基于摘要

提出静态分析方法识别 React 代码中的 Prop Drilling,并用 Context API 与组件组合两种自动化重构策略消除它。

cs.SE

查看原文摘要

In React front-end development, Prop Drilling is the practice of propagating data through component properties across multiple levels of the component hierarchy. Despite being discouraged by React documentation and characterized as a code smell by recent research, little is known about its prevalence, complexity, and potential for automated refactoring. In this work, we propose a static analysis method for identifying Prop Drilling instances in a React codebase and eliminating them using two automated refactoring strategies based on Context API and Component Composition. The proposed method is implemented as a Nodejs command line tool, ReactRefactor, and empirically evaluated on a benchmark dataset of open-source React applications. The main findings of the empirical evaluation indicate (a) the frequent occurrence of prop drillings across benchmark projects, irrespective of project size; (b) their generally low to moderate complexity in terms of data propagation depth, contrasted with the more prevalent complexity arising from the simultaneous forwarding of multiple properties along the same path; and, (c) the potential for automated elimination of a substantial share (76.3%) of prop drillings, primarily through refactoring to the Context API.

版本更新 · v3arXiv:2603.24562

面向临床记录的可扩展复发感知基础模型:基于下次就诊预测

Scaling Recurrence-aware Foundation Models for Clinical Records via Next-Visit Prediction

Xiang Gao, Haresh Rengaraj Rajamohan, Weicheng Zhu, Shih-Lun Huang, Long Chen, Gabe Schulman, Huizhen Jin, Shengduo Li 等 13 位作者

AI 速读 基于摘要

提出 RAVEN,一种基于复发感知下次就诊事件预测的序列 EHR 生成式预训练策略,并在百万级个体数据上考察其零样本预测与缩放行为。

cs.LG

查看原文摘要

While large-scale pretraining has revolutionized language modeling, its potential remains underexplored in healthcare with structured electronic health records (EHRs). We present RAVEN, a novel generative pretraining strategy for sequential EHR data based on Recurrence-Aware next-Visit EveNt prediction. Leveraging a dataset of over one million unique individuals, our model learns to autoregressively generate tokenized clinical events for the next visit conditioned on patient history. We introduce regularization on predicting repeated events and highlight a key pitfall in EHR-based foundation model evaluations: repeated event tokens can inflate performance metrics when new onsets are not distinguished from subsequent occurrences. Furthermore, we empirically investigate the scaling behaviors in a data-constrained, compute-saturated regime, showing that simply increasing model size is suboptimal without commensurate increases in data volume. We evaluate our model via zero-shot prediction for forecasting the incidence of a diverse set of diseases, where it rivals fully fine-tuned representation-based Transformer models and outperforms both standard simulation-based next-token approaches and a prompted medical large language model baseline. Finally, without additional parameter updates, we show that RAVEN can generalize to an external patient cohort under lossy clinical code mappings and feature coverage gaps.

版本更新 · v2arXiv:2607.06724

EvoPlan: Evolutionary Neuro-Symbolic Robot Planning with Spatio-Temporal Guarantees

Bhavya Sai Nukapotula, Samin Moosavi, Haoze Wang, Luke Duncan, Diya Shakkottai, Varun Murali, Srinivas Shakkottai

原文摘要

LLM-based robot planners are fluent but cannot guarantee that their plans are executable or safe. Classical PDDL planners can guarantee symbolic validity, but only once the problem is fully specified, and they make poor use of an LLM's ability to read context and repair plans. We present EvoPlan, a neuro-symbolic framework whose guarantees are symbolic validity and satisfaction of a constraint learned from data. The planner works with locally-hosted open-weight models, so it does not require a cloud service. EvoPlan has three parts. First, an offline procedure that mines a single global Signal Temporal Logic (STL) mobility constraint from demonstrations. Perturbed demonstrations supply the missing negatives without encoding any rule, and evolutionary search fits the constraint, which captures driving rules or social-navigation preferences within a modelled taxonomy of norms. We use the mined constraint to shield driving and navigation policies on Bench2Drive and HA-VLN-CE. Second, an evolutionary PDDL planner: an LLM proposes and repairs plans, programmatic validators decide which ones survive, and the validated portion of the plan grows over iterations. On open-world ALFWorld and NL-PDDL Blocksworld, it approaches classical planning when the goal vocabulary agrees with the action model, and it stays robust when the two disagree and classical planning fails. Third, a constrained execution loop: Nav2 uses the costmap to ground the plan into waypoints, which are checked against the mined constraint. Violations and execution failures update the PDDL problem for re-planning. We demonstrate the full pipeline in Gazebo and on a Clearpath Jackal robot. More details and videos are on the project page: https://nbhavyasai.github.io/evoplan/

cs.RO

版本更新 · v2arXiv:2610.37973

合并与分支:面向推荐的大语言模型高效树结构监督微调

Merge and Branch: Efficient Tree-Structured Supervised Fine-Tuning of LLMs for Recommendation

Yunqin Zhu, Liyang Shu, Yusi Zhang, Yao Xie, Tianhao Li, Xinhang Song, Tianyu Wen

AI 速读 基于摘要

提出 Merge and Branch(M&B)方法,通过压缩历史表示并用树结构共享历史计算、同时监督多个目标,提升基于 LLM 的生成式推荐的训练与解码效率。

cs.IR · cs.AI

查看原文摘要

LLM-based generative recommendation reads long, text-rich interaction histories but generates only short semantic identifiers (SIDs). This asymmetry makes supervised fine-tuning inefficient: encoding the history is expensive, yet each history usually provides supervision for only one target. We present Merge and Branch (M&B) to reduce input cost and learn more from each encoded history. Merge compresses each historical item's SID and text into a small number of learned tokens. Branch uses the compressed history to predict multiple target SIDs. To share history computation while keeping these targets independent, M&B organizes each example as a shallow, wide tree. We develop tree FlashAttention to execute this layout efficiently, processing the shared history once and propagating gradients from every branch for end-to-end training. A tree-structured key/value cache extends history sharing to constrained decoding. Experiments on public benchmarks show that M&B outperforms strong recommendation baselines while substantially improving training and decoding efficiency. Online A/B tests on a large-scale e-commerce platform suggest that M&B effectively combines item text with multi-target supervision, yielding statistically significant gains in per-user clicks, orders, and GMV.

版本更新 · v2arXiv:2608.22154

混合贝叶斯-联结主义模型实现更准确的行为预测

More accurate behavioral predictions with hybrid Bayesian-connectionist models

Brenden M. Lake, Akshay K. Jagadish, Guangyuan Jiang

AI 速读 基于摘要

作者提出贝叶斯蒸馏与行为微调(BBT)方法,将贝叶斯模型的可解释先验与神经网络对复杂行为的拟合能力结合,并在人类概念学习的四个案例中报告了更优的行为预测表现。

cs.LG

查看原文摘要

Researchers must often choose between Bayesian or neural network models of behavior, two paradigms with complementary strengths and weaknesses. An ideal paradigm would facilitate testing many kinds of representations and inductive biases; Bayesian models make this easy, while neural networks do not. Similarly, an ideal paradigm would avoid over-simplifications; neural networks make this easy, while Bayesian models do not. Here, we introduce Bayesian distillation with Behavioral Tuning (BBT) as an approach to getting the best of both traditions. BBT offers a simple recipe for model building: first, a neural network is trained to mimic a Bayesian model through synthetic data, and second, the network is fine-tuned on human behavior to capture additional structure and nuance. Across four case studies in human concept learning, we find that BBT outperforms traditional approaches at predicting human behavior while also revealing psychological insights, resulting in models that can both mimic Bayesian priors and capture heuristics and biases that violate simple modeling assumptions.

版本更新 · v2arXiv:2610.31661

OpenTac:面向可定制、可扩展与可迁移三维触觉感知的开放生态

OpenTac: An Open Ecosystem for Customizable, Scalable, and Transferable 3D Tactile Sensing

Ariel Slepyan, Geeta Chandra Raju Bethala, Jiafeng Gu, Anagha Rajendra Dangle, Yinzhe Zhou, Hongyuan Fu, Jade Sim, Tuhin Basu 等 15 位作者

AI 速读 基于摘要

作者提出一种可扩展的电阻式三维触觉皮肤及配套设计工具、仿真器与数据采集生态,并在多种机器人平台上演示其用于控制与策略学习。

cs.RO · eess.SP · eess.SY

查看原文摘要

We present OpenTac, a novel sensor and a full-stack ecosystem for tactile skins that are highly customizable, scalable, and readily deployed in downstream robot control and learning applications. Our core contribution is twofold. First, we develop a scalable resistive tactile skin with a structured deformable overlay that converts tangential loading into spatially varying pressure. Differential responses across neighboring "taxels" support decoding of local shear vectors, and the common-mode response encodes normal force. Second, we provide a highly customizable sensor design tool that is able to automatically generate fabrication files for these tactile skins from 2D outlines and developable 3D surfaces, including from abstract descriptions such as text. We develop an ecosystem around the OpenTac sensor, including the CAD design tool, a sensor simulator, cross-sensor transfer, and fast/adaptive data acquisition. We demonstrate the broad applicability of the OpenTac-CAD sensor design tool by designing and fabricating more than 30 sensors across a variety of robot embodiments. We characterize the three-dimensional tactile fields produced by OpenTac with a commercial, high-precision force/torque sensor. We demonstrate the broad applicability of 3D tactile sensing across contact-rich feedback control and manipulation policy learning on four real-world robot embodiments. Additionally, adaptive acquisition enables kilohertz-rate partial readouts, and our readout circuitry enables high dynamic range (HDR) acquisition, measuring, for example, a 3.7 kg weight stack and a 20 g weight on the same sensor in the same frame. We provide all hardware, software, and result videos at opentac.github.io

版本更新 · v2arXiv:2606.21036

扩散驱动状态空间模型

Diffusion-Driven State Space Models

Jack Ruder, Michael Wojnowicz

AI 速读 基于摘要

用扩散模型替代高斯转移分布构建深度状态空间模型,并解决序列数据上自编码器与扩散模型的联合训练问题。

stat.ML · cs.LG

查看原文摘要

In many domains, practitioners seek models that produce accurate forecasts while faithfully capturing latent system dynamics. Existing approaches typically sacrifice one of these goals: deep state space models often assume Gaussian latent transitions, limiting fit and forecasting, while diffusion models are highly expressive but lack principled inference for the underlying dynamics. To combine the strengths of both, we introduce the Diffusion-Driven State Space Model (DDSSM), which replaces the conventional Gaussian transition distribution with a diffusion model. Our DDSSM resolves the open problem of how to jointly train an autoencoder and a diffusion model on sequential data, thereby extending the literature on latent diffusion models for time series. Moreover, we find that the DDSSM empirically outperforms a state-of-the-art deep SSM at fitting and forecasting a simulated time series with multimodal transitions.

版本更新 · v3arXiv:2505.22941

唯一的 Class 0 Flower snark 是最小的那个

The only Class 0 Flower snark is the smallest

Guilherme Adamatti Bridi, André Luis Alves Martins, Franklin de Lima Marquezino, Celina Miraglia Herrera de Figueiredo

AI 速读 基于摘要

作者证明最小的 Flower snark J_3 是 Class 0,从而说明 J_3 是唯一为 Class 0 的 Flower snark。

math.CO · cs.DM

查看原文摘要

In graph pebbling, pebbles are distributed across the vertices of a graph. A pebbling move consists of removing two pebbles from one vertex and adding one pebble to an adjacent vertex. The pebbling number is the minimum number $t$ such that, regardless of how the $t$ pebbles are initially placed, it is always possible to perform a sequence of moves that results in at least one pebble placed on any specified target vertex. Graphs whose pebbling number is equal to the number of vertices are called Class~$0$ and provide a challenging set of graphs that resist being characterized. In this note, we answer a question recently proposed by the pioneering study on the pebbling number of snark graphs: we prove that the smallest Flower snark $J_3$ is Class~$0$, establishing that $J_3$ in fact is the only Class~$0$ Flower snark. that systematically reduce the space of pebble configurations. Our bound-based approach shows particular promise for diameter $3$ graphs, where the neighborhood structure remains sufficiently constrained to enable effective analysis.

版本更新 · v3arXiv:2601.05416

头部何时离开?360度视频观看中的共享时序与运动时钟

When does the head leave? Shared timing and a movement clock in 360-degree video viewing

Arman Nik Khah, Arvin Bahreini, Ravi Prakash

AI 速读 基于摘要

基于四个公开头部追踪数据集,作者报告360度视频观看中同区域观众存在秒级同步离开现象,且距上次大幅头部运动的时间可提升离开预测,眼动追踪测试集上 gaze 滞后于头部启动约0.30秒。

cs.MM

查看原文摘要

A $360^\circ$ streaming system only needs high quality where the viewer faces, so knowing when a head will turn away matters as much as where it will go. Across four public head-tracking datasets, two used to fix and validate the method and two held out for testing, three patterns emerge on the held-out sets. First, viewers facing the same region turn away within a second of one another far more often than time-shifted traces allow. That estimate is about five times the pairwise threshold for a video timing one departure in ten, mapping to roughly one and a third to over twice that one-in-ten share. The excess co-departing fraction is 0.0422, with a 90% interval from 0.0279 to 0.0565. Second, time since the last large head movement improves held-out departure prediction beyond video and viewer effects. It adds 0.0793 nats per departure, a predictive dependence rather than proof of an internal timer. Third, on the eye-tracked test set, gaze crosses the region boundary after the head starts moving, lagging head onset by 0.30 s. Because headsets record continuous angles, a departure exists only after a rule marks when a settled hold ends once the head swings past a speed and angle threshold. Before touching the test sets, we picked one primary setting for the headline estimates above and tested 18 settings on synthetic traces with planted answers, requiring at least nine passing settings for robustness. Fifteen settings passed for the clock and all 15 valid settings confirm its gain; 8 passed for synchrony and all 8 valid settings agree, one short of the nine-setting bar; and 6 of 7 dual-validated settings agree on gaze lag. Synchrony validation drops from 17 or 18 passing settings under guessed noise to 8 under measured noise because zero-tolerance rules reject a small baseline offset, consistently positive at one dataset size, as datasets grow.

版本更新 · v2arXiv:2611.06476

对齐税是分级的:长度归一化DPO结合SFT保留项下的任务选择性能力损失

The Alignment Tax Is Graded: Task-Selective Capability Loss Under Length-Normalized DPO with SFT Retention

Ye Yuan

AI 速读 基于摘要

作者报告,在长度归一化DPO加SFT保留项(RPO目标)下,对齐税呈分级且任务选择性:抽取能力基本保持,json小幅下降,router严重退化。

cs.LG

查看原文摘要

Setting. This is an empirical characterization study; we propose no new method and run no comparisons against alternative objectives. We fine-tune Qwen3.5-4B with sparse LoRA and length-normalized DPO plus an SFT retention term (the RPO objective) on structured financial-filing tasks, and track downstream capability as the implicit reward margin grows to $\sim 7$ $β$-scaled units (nf4 training; $\sim 3$ under fp16 measurement on the unquantized base). Question. Does margin growth degrade capability uniformly across tasks, or selectively? Findings. In a pre-registered study (8 seeds, 6 unseen at freeze) with artifact-aware scoring and TOST equivalence testing, the alignment tax is graded and task-selective: extraction is preserved ($Δ= -0.55$ pts, equivalent at $\pm5$ pts), json loses little ($Δ= -3.52$ pts, equivalence fails on a clustered item-level interval), and router degrades severely ($Δ= -30$ pts). Router's loss concentrates in a margin 1-3 damage window containing a transient format-break spike. The margin is accounted for almost entirely ($\geq 96\%$) by suppression of rejected completions, with chosen reward approximately flat, consistent with the retention term anchoring it; we do not attribute the behavioral damage to any single gradient channel. Gold-logp deflation is task-selective (extraction $-0.15$, json $-1.04$, router $-1.47$ nats/token under fp16), with gold selectively protected relative to incorrect controls on extraction (8/8 seeds). QLoRA reward margins differ up to $3.4\times$ between nf4 training and unquantized fp16 measurement, with an axis-dependent gap. Scope. Length-normalized rewards with SFT retention, sparse LoRA on 8/32 layers, one 4B model, easily-separable synthetic preferences, three structured tasks. A no-retention control and a measured format-alignment metric are left to future work.

版本更新 · v2arXiv:2611.05612

Whiel:面向可认证关系程序验证的神经符号平台

Whiel: A Neurosymbolic Platform for Certified Relational Program Verification

Jesse Comer, Fangzhu Shen, Shuo Zhang, Mayur Naik, Sudeepa Roy, Val Tannen

AI 速读 基于摘要

Whiel 是一个用 Lean 证书认证命令式关系程序推理的神经符号系统,支持 Datalog 及其否定扩展、图查询核心片段与数据集成规格的包含/等价问题。

cs.DB · cs.AI · cs.LO

查看原文摘要

We present the Whiel (a contraction of Whi(le-R)el) system for Lean-certified reasoning about imperative relational programs. The system can encode containment and equivalence problems for Datalog, its extensions with negation, and the core fragment of graph query languages. It can also represent optimized or incremental evaluation algorithms that are not expressible in the original query language, as well as first-order specifications of integrity constraints, schema mappings, and views arising in data integration, exchange, and sharing. Whiel's program states are database instances, and its imperative structure enables compositional Hoare-style reasoning with first-order relational specifications, using prophecy variables and leveled invariants to support inductive arguments. Whiel supports neurosymbolic discovery of loop invariants whose verification conditions can be checked using the Vampire first-order theorem prover. Counterexamples are fed back into the neurosymbolic pipeline, while proposed solutions are translated into Lean proof certificates. To validate the Whiel system, we evaluate its performance on a curated benchmark suite. In summary, we envision the Whiel system as a tool for tackling mathematically demanding, typically undecidable problems involving complex query and constraint languages, with a high degree of trust in its output.

版本更新 · v5arXiv:2610.26261

REFLEX: RNA Ensemble Generation via Flexibility-Calibrated Stochastic Bridge

Wanli Ma, Tianmeng Hu, Ke Li

原文摘要

RNA molecules often function through multiple experimentally observed conformational states. A central challenge is to capture ensemble diversity without sacrificing structural fidelity: samples may collapse toward averaged representative structures or drift into unsupported conformational regions. We introduce REFLEX (RNA Ensemble generation calibrated by FLEXibility), a sequence-conditioned generator for RNA backbone-frame ensembles based on a heteroscedastic stochastic bridge on SE(3). During training, experimental B-factors calibrate residue-wise bridge widths, exposing each observed conformer through local perturbation neighborhoods with residue-specific scales while keeping constrained regions sharper. A learned flexibility condition guides the bridge-field network to recover target conformer-specific frames from these calibrated noisy states. At inference, REFLEX uses only the input sequence, requiring no MSA-subsampling heuristics. On a multi-conformer RNA benchmark, REFLEX achieves the best precision-recall trade-off for ensemble coverage across evaluation thresholds, while recovery-stress ablations indicate gains in target-frame recovery under stronger perturbations.

cs.AI

版本更新 · v6arXiv:2305.04281

用持续同调分析多尺度聚类

Analysing Multiscale Clusterings with Persistent Homology

Juni Schindler, Mauricio Barahona

AI 速读 基于摘要

作者提出多尺度聚类滤流(MCF),用拓扑数据分析把任意多尺度划分序列编码为单纯复形滤流,并用持续图刻画跨尺度的不一致性。

math.AT · cs.LG

查看原文摘要

In data clustering, it is often desirable to find not just a single partition into clusters but a sequence of partitions that describes the data at different scales (or levels of coarseness). A natural problem then is to analyse and compare, as a whole, the (not necessarily hierarchical) sequences of partitions that underpin such multiscale descriptions. Using tools from topological data analysis, we introduce here the Multiscale Clustering Filtration (MCF), a filtration of abstract simplicial complexes that encodes arbitrary cluster assignments in a sequence of partitions across scales of increasing coarseness. We prove the stability of the MCF with respect to small changes in the sequence of partitions, and establish that its persistence diagram (PD) provides an interpretable representation of the multiscale clustering: at dimension 0, it captures violations of the refinement order, and at higher dimensions, it tracks the emergence and resolution of higher-order inconsistencies between cluster assignments. We also introduce an equivalent construction of the MCF via a nerve complex filtration, and show that this nerve-based MCF has computational advantages in certain settings, aiding scalability. Further, we show that in the hierarchical case, the MCF reduces to a Vietoris-Rips filtration of an ultrametric space, and we establish that pairwise measures, such as conditional entropy or zigzag persistence, are fully determined by a graph representation of the Sankey diagram and thus fail to describe the higher-order inconsistencies across scales that MCF captures. Using synthetic and real-world data from stochastic block models, single-cell transcriptomics and temporal social interactions, we illustrate how the MCF PD serves to characterise and compare multiscale clusterings beyond pairwise or hierarchical baseline methods.

版本更新 · v3arXiv:2610.11093

风险上限与发展截止期:在安全生产力不确定下为AI发展定节奏

Risk Ceilings and Development Deadlines: Pacing AI under Uncertain Safety Productivity

Li Gan

AI 速读 基于摘要

研究在安全生产力未知时,监管者能否同时承诺风险上限与发展截止期,并探讨先学习后重放开发、预留安全算力的策略。

econ.GN · cs.CY

查看原文摘要

Can a regulator promise both a risk ceiling and a development deadline when safety productivity is unknown? A ceiling below the final model's unprotected hazard requires a minimum stock of safety knowledge, so both promises hold only if weak research can be ruled out. Learning first and then replaying development, with spare compute in safety, comes close to that minimum. In a calibration with only state risk, constant safety yield, and full knowledge transfer, it needs only 3.3 percent more productivity than the necessary bound. Customer services pay for the guarantee. Rules that fix compute allocation fix dates, not risk.

版本更新 · v2arXiv:2610.01172

混合Transformer-SSM架构的学习率迁移

Learning Rate Transfer for Hybrid Transformer-SSM Architectures

Jimin Seo, Gyubok Lee, Yeonsik Jo, Kiwoong Yoo, Yeongoon Kim, Minhae Oh, Jin Woo Koo, Suhwan Kim 等 14 位作者

AI 速读 基于摘要

作者报告在混合Transformer-SSM架构中,仅用原始μP处方即可实现跨宽度和深度的近零学习率迁移差距,并归因于全局更新-权重不变性与AdamW局部平衡。

cs.LG

查看原文摘要

We study learning rate (LR) scaling for hybrid architectures combining Transformer and State-Space Model (SSM) blocks, a class adopted by several recent production language models. In particular, we focus on the gap between the theoretical scaling rules derived for SSMs under zero-order-hold (ZOH) discretization at infinite width with growing state size, and the field-standard practical implementations using simplified-ZOH Mamba at fixed state size. Surprisingly, in this practical regime hybrid architectures achieve a near-zero LR transfer gap across widths 256-2048 and depths 4-32 up to billion-parameter scale using only the original $μ$P prescription, even though SSM operations fall outside its Tensor Programs representability conditions and every parameterization we test fails the standard coordinate-check diagnostic of $μ$P correctness. We attribute this to a two-condition decomposition of LR transfer in hybrid architectures: a global update-to-weight invariance, enforced by $μ$P's initialization and LR scaling; and a local per-component balance, provided by AdamW's per-parameter normalization. Our observations show that the optimal LR is invariant to width up to 8$\times$, that this width invariance holds across depth, sequence length, batch size, and Transformer-to-SSM ratio, and that it transfers to Nemotron-H, a production hybrid outside our custom architecture set. We hope these findings fill the gap between theoretical scaling rules and practical hybrid implementations, and stimulate further research toward bridging it.

版本更新 · v3arXiv:2605.12850

涌现性失配中的人格模型坍缩

Persona-Model Collapse in Emergent Misalignment

Davi Bastos Costa, Renato Vicente

AI 速读 基于摘要

作者提出并检验“人格模型坍缩”假说:有害数据微调导致模型内部人格表征退化,表现为角色扮演回答更极化、一致性下降且激活维度收缩。

cs.CL · cs.AI · cs.CR · cs.LG

查看原文摘要

Fine-tuning large language models on narrow data with harmful content produces broadly misaligned behavior on unrelated prompts, a phenomenon known as emergent misalignment. We propose that emergent misalignment involves persona-model collapse: a deterioration of the internal representations that encode and distinguish personas, undermining the nuance and consistency of their expression. We first test this hypothesis behaviorally, by comparing how models fine-tuned on harmful data and benign controls answer questionnaires while role-playing diverse personas. Across four model families, harmful fine-tuning produces more polarized, all-or-nothing answers, suggesting more binary, less nuanced persona expression. Histograms show increased concentration at the scale endpoints, and Kullback--Leibler (KL) divergence quantifies the distributional shift. We also find reduced within-persona consistency: repeated answers to the same question under the same persona become less stable. We reproduce these observations across diverse harmful datasets and questionnaires that probe different aspects of persona-conditioned behavior, while capability and simple instruction-following controls argue against a uniform loss of general ability. Finally, a mechanistic analysis of persona-conditioned activations in Qwen3-235B shows that representations elicited by different persona prompts span fewer effective dimensions in the misaligned variant than in the original and benign-control variants, most clearly in the final layers. Together, these findings provide behavioral and representational support for persona-model collapse as an aspect of emergent misalignment.

版本更新 · v2arXiv:2609.30467

基于检索的开放式评估在哪里失败?从长篇医学答案事实性验证中自动归纳分类体系

Where Does Retrieval-Based Open-Ended Evaluation Fail? Automatic Taxonomy Induction from Long-Form Medical Answer Factuality Verification

Heyuan Huang, Jirui Dai, Alexandra DeLucia, Sonal Joshi, Mahsa Yarmohammadi, Jie Gao, Bernal Jiménez Gutiérrez, Mark Dredze

AI 速读 基于摘要

作者针对开放式医学答案的检索式事实性验证,构建两套失败分类体系,并用自动模式归纳与压力测试揭示检索与验证环节的系统性失败模式。

cs.CL · cs.IR

查看原文摘要

Retrieval-based factuality evaluation, where LLM-generated claims are verified against evidence from authoritative medical corpora, has become the dominant paradigm for scalable hallucination detection in high-stakes clinical settings. Despite the urgency of reliable and transparent medical fact verification, most systems measure performance with aggregate metrics like F1, which obscure where and why failures occur. Existing RAG diagnostics require gold answers or annotated gold evidence, neither of which exists in this regime. We introduce two comprehensive taxonomies, grounded in a case study on the open-ended MedExpert dataset and 3 closed-ended datasets, decomposing failures into retrieval-stage errors along five quality dimensions, and verifier-reasoning errors into six consecutive steps. We adapt an automatic pattern induction pipeline using LLM-as-Judge to label evidence quality and classify verifier reasoning errors at scale, and then stress-test our findings across 4 retrieval methods and 6 frontier verifier models. Our analysis reveals that scaling model size, adding reasoning effort, expanding to authoritative web sources, and applying medical fine-tuning do not resolve these failure modes, demonstrating that they represent fundamental limitations of the retrieve-then-verify paradigm in open-ended medical settings rather than artifacts of outdated systems. We release our code and data at https://anonymous.4open.science/r/Medical_RAG_eval-4AB5 for the full reproducibility of our results.

版本更新 · v2arXiv:2605.27354

按模型思考方式教学:用稀疏自编码器指导LLM后训练数据工程

Teach the Way It Thinks: Guiding LLM Post-training Data Engineering with Sparse Autoencoders

Yi Jing, Zao Dai, Jinwu Hu, Zijun Yao, Lei Hou, Juanzi Li, Xiaozhi Wang

AI 速读 基于摘要

提出SAERL,利用稀疏自编码器(SAE)特征空间中的模型内部信号来组织强化学习训练数据的选择、排序与批处理。

cs.LG · cs.AI · cs.CL

查看原文摘要

Data engineering for large language model (LLM) reinforcement learning (RL) decides which samples to train on, when to present them, and which to batch together. Existing methods mostly rely on external feedback, such as pass rates or difficulty labels, which observes only the outcome on each sample. We argue that model internals offer richer and more model-intrinsic feedback that can guide all three decisions. We propose SAERL which organizes RL training data in the feature space of Sparse Autoencoders (SAEs), a mechanistic interpretability tool that decomposes model internals into sparse, interpretable features. SAE features let us view the model as a set of functional regions, each formed by features that related samples activate together. For each sample, its SAE activations indicate which regions it engages, and its gradients projected onto SAE features indicate how training on it would change these regions. Guided by this view, SAERL keeps samples with strong and non-redundant gradient signals, trains each functional region from easy to hard, and trains one functional region at a time, drawing each batch from a single activation cluster with moderate cross-cluster mixing. The resulting curriculum is constructed once with small models and transfers to larger ones. Experiments on mathematical reasoning show consistent gains across model scales and RL algorithms. For instance, on Qwen2.5-Math-1.5B, SAERL improves the average accuracy of GRPO by 3.1% and reaches the target accuracy with 42% fewer training steps. Further analyses reveal that ordering and batching are complementary. These results demonstrate that model internals are a powerful and practical foundation for post-training data engineering.

版本更新 · v2arXiv:2606.28273

视觉默认、先验覆盖:视觉语言模型中感知—知识冲突的因果机制

Vision-Default, Prior-Override: Causal Mechanisms of Perception-Knowledge Conflict in Vision-Language Models

Niclas Lietzow, Danielle Bitterman, Carsten Eickhoff, William Rudman, Michal Golovanevsky

AI 速读 基于摘要

该研究用激活修补与组件消融揭示:视觉语言模型默认依赖视觉证据,而先验知识覆盖视觉输入依赖少量因果必要的注意力头。

cs.CL

查看原文摘要

Vision-language models must reconcile visual evidence with memorized world knowledge when the two conflict. How they resolve this conflict shapes the reliability of multimodal systems, yet prior work characterizes it behaviorally without a component-level causal account. We combine activation patching across three granularities (residual stream, attention heads, and MLP sublayers) with model-component ablation studies and mechanistic analysis. Across three VLM families, we find that visual grounding emerges by default, whereas prior grounding depends on a small set of causally necessary attention heads (2.5-4.8%) concentrated in the second half of the network. These heads enable answers from stored world knowledge (e.g., "red" for a strawberry) despite conflicting visual input. Ablating them flips predictions from knowledge-grounded to visually grounded answers in 68-96% of cases under prior-knowledge prompts, but changes only 0.8-7.5% of visually grounded predictions, establishing an asymmetric causal structure. The identified heads decompose into routing heads, which modulate information flow, and writing heads, which directly project answer tokens into the residual stream. This structure is consistent across model families and scales, revealing a sparse causal circuit underlying perception-knowledge conflict in VLMs.

版本更新 · v2arXiv:2605.17184

显著、可分解且不可见:物理任务教学视频中的视觉情境错位

Substantial, Decomposable, and Invisible: Visual Context Misalignment in Instructional Videos for Physical Tasks

Yayuan Li, Chenglin Li, Jingying Wang, Filippos Bellos, Anhong Guo, Jason J. Corso

AI 速读 基于摘要

作者通过两项研究考察教学视频与用户视觉情境不匹配的影响,发现对齐能提升完成质量与速度,且该效应可分解为四类视觉情境属性,但用户对单属性错位造成的影响缺乏感知。

cs.HC

查看原文摘要

Instructional videos are the dominant medium for completing physical tasks, yet rarely match the user's visual context. Motor simulation and cognitive load theories predict this mismatch matters, but how much, which attributes are responsible, and how users experience it remain unknown. We conduct two studies (56 participants, two first-aid and two cooking tasks) using researcher-prepared recordings to control visual alignment. Study 1 (N=16) compares fully aligned In-Context instructional videos (ICON) against business-as-usual Internet videos. ICON yields statistically significant gains: 11.1% higher completion quality and 15.5% faster completion. Qualitative analysis decomposes the effect into four visual context attributes: Task Object Intrinsic, Task Object State, Environmental Context, and Observational Context. Study 2 (N=40) ablates attributes one at a time. All four degrade objective performance yet users perceive no performance loss for any of them. Visual context misalignment is substantial, decomposable, and invisible to the user for any single attribute.

版本更新 · v4arXiv:2510.22386

关于 R^3 中平移盒体自由空间复杂度

On the complexity of the free space of a translating box in R^3

Gabriel Nivasch

AI 速读 基于摘要

作者证明在 R^3 中平移盒体(以及边成平行对的凸多边形)的自由空间组合复杂度为 O(n^2),从而解决盒体及所有凸多边形的渐近最坏情况复杂度问题。

cs.CG

查看原文摘要

Consider a convex polyhedral robot $B$ that can translate (without rotating) amidst a finite set of non-moving polyhedral obstacles in $\mathbb R^3$. The "free space" $\mathcal F$ of $B$ is the set of all positions in which $B$ is disjoint from the interior of every obstacle. Aronov and Sharir (1997) derived an upper bound of $O(n^2\log n)$ for the combinatorial complexity (number of vertices, edges, and faces) of $\mathcal F$, where $n$ is the total complexity of the obstacles, and the complexity of $B$ is assumed constant. Halperin and Yap (1993) showed that, if $B$ is either a box or a "flat" convex polygon, then a tighter bound of $O(n^2α(n))$ holds. Here $α(n)$ is the inverse Ackermann function. In this paper we prove that if $B$ is a box, then the complexity of $\mathcal F$ is $O(n^2)$. Furthermore, if $B$ is a convex polygon whose edges come in parallel pairs, then the complexity of $\mathcal F$ is $O(n^2)$ as well. These results settle the question of the asymptotical worst-case complexity of $\mathcal F$ for a box, as well as for all convex polygons.

版本更新 · v2arXiv:2610.07274

面向智能体评估的可信层

A Trust Layer for Agent Evaluation

Mohammadreza Sediqin, Shivali Dalmia, Srinivasa Karthikeya Reddy Kovvuri, Abhishek Mukherji

AI 速读 基于摘要

提出一种附加式事后框架,在记录分数旁标注该分数是否可信,并用四项检查验证智能体评估结果。

cs.AI

查看原文摘要

Deterministic benchmark scores show that an agent received credit, but not whether that credit was earned, reported honestly, or would hold on a second run. We introduce a Trust Layer for Agent Evaluation, an additive post-hoc framework that reports, beside each recorded score, whether it should be believed. It verifies four properties: whether the result is supported by the benchmark's own grading logic, whether a passing answer was earned through traceable computation, whether the agent's completion claim matches what occurred, and whether the result is stable under repeated execution. The first three use only saved artifacts; the fourth re-runs the agent. Model judgments only label evidence under majority voting; all verdicts follow deterministic rules and never modify the recorded score. Applied to five agent configurations on 108 tasks from Agents' Last Exam, every model shows passing runs with no traceable computation (at rates varying tenfold), confirmed false completion claims, and unstable results: 18-46% of tasks do not stay in one score band over five runs. Only 22.6% of recorded passes clear all four checks (95% CI 15.0-32.6, n=84). Measuring what an agent can do and verifying that it did it are different problems, and current benchmarks address only the first.

版本更新 · v2arXiv:2606.23431

DexTeleop-0: Force-Aware Bimanual Dexterous Teleoperation with Ego-Centric Perception towards Shared Autonomy

Haichao Liu, Yuyao Jiang, Hyunsun Park, Yuanjiang Xue, Mingze Sun, Ziwei Liu, Ziwei Wang

原文摘要

Fine-grained, bimanual dexterous manipulation remains a foundational challenge in robotics. Embodiment gaps hinder accurate kinematic mapping, and visual tracking alone does not regulate contact forces. To address these limitations, we propose a tactile-driven adaptation strategy designed to enable fine-grained manipulation on top of teleoperation pipelines. Instantiated within our bimanual dexterous framework, DexTeleop-0, this strategy introduces a real-time optimization loop that adds robot-side contact corrections to human tracking commands using fingertip feedback. By estimating contact positions and forces, the system dynamically computes localized corrections using a contact-stiffness approximation and the kinematic Jacobian, with per-contact vector force tracking, soft predicted-contact-moment regularization, and bounded hand-joint residuals. We evaluate this tactile-driven adaptation strategy across both simulated environments and real-world hardware. Controlled simulation replay yields final-stage success rates of 91% for Ball Assembly and 62% for Stir in Cup, versus 88% and 49% for force tracking alone. Fixed-input replay on two hand types, seven-operator teleoperation, and external-force traces further assess contact assistance. The open-sourced project page can be found at https://henryhcliu.github.io/dexteleop-0.

cs.RO

版本更新 · v3arXiv:2512.15596

纠错扩散语言模型

Corrective Diffusion Language Models

Shuibai Zhang, Fred Zhangzhi Peng, Yiheng Zhang, Jin Pan, Grigorios G. Chrysos

AI 速读 基于摘要

作者提出一种面向纠错的扩散语言模型后训练目标,让模型能对可见的错误 token 赋予较低置信度并迭代修正,同时保留正确内容。

cs.LG

查看原文摘要

While Diffusion Language Models (DLMs) are theoretically well-suited for iterative refinement due to their non-causal structure, off-the-shelf masked diffusion language models (MDLMs) often fail to reliably revise incorrect tokens in practice. The key challenge is that the model cannot tell correct from erroneous visible tokens: standard MDLM training supervises only masked positions, undermining confidence-guided refinement. We study corrective behavior in DLMs, the ability to assign lower confidence to incorrect tokens and iteratively refine them while preserving correct content. Under matched compute and data, we show that conventional masked-diffusion fine-tuning induces much less of this capability than a correction-oriented post-training principle that explicitly supervises visible corrupted tokens, which enables discriminative confidence and targeted refinement. We further find that also supervising clean visible tokens, as in BERT- or GIDD-style training, stops remask-based refinement from improving with additional steps. To evaluate it, we introduce the Code Revision Benchmark, a controllable, executable benchmark for error localization and in-place correction. The effect transfers from a 0.5B model to LLaDA-8B and persists, with a smaller margin, when corruptions are drawn from the base model's own most plausible predictions. The same error-aware confidence also makes parallel decoding more effective. In HumanEval(+) code completion at 0.5B, mathematical reasoning at 16B, and high-entropy parallel decoding with remasking (ParallelBench), models trained with our objective outperform standard masked-diffusion fine-tuning under matched compute, and outperform the base model on HumanEval(+), MATH, and the high-entropy ParallelBench average; at 8B, neither objective improves from-scratch generation over the base model. Code and data: https://github.com/zhangshuibai/CDLM

版本更新 · v2arXiv:2605.23504

VACE:为时间序列异常检测学习几何结构化表示

VACE: Learning Geometrically Structured Representations for Time Series Anomaly Detection

Alberto D. Cencillo, Leonardo Concepción, Isaac Triguero, Julián Luengo

AI 速读 基于摘要

VACE 通过速度一致性目标训练通道感知编码器,把正常多元时间序列表示为嵌入空间中局部平滑、方向连贯的轨迹,并用位置与方向得分相乘进行异常检测。

cs.LG · cs.AI

查看原文摘要

Anomaly detection in multivariate time series is a critical task across a wide range of real-world applications, where abnormal behaviour is rare, labels are unavailable, and the cost of a miss is high. The central challenge is learning a characterisation of normality precise enough to flag deviations. Representation self-supervised learning, typically through contrastive approaches, addresses this by embedding temporal patches into a latent space where normality occupies a well-defined region, with anomalies detected by geometric deviation. However, contrastive approaches shape this space indirectly through pair-sampling heuristics, providing no explicit control over the geometric structure on which distance-based scoring operates. That geometry is thus left to chance. We present VACE (Velocity-Aligned Channel Embeddings), a self-supervised anomaly detection method that represents normality as a locally smooth, directionally coherent trajectory in the embedding space. To this end, VACE trains a channel-aware encoder through a velocity-consistency objective, with no negatives and no synthetic anomalies, so that normal trajectories are locally smooth and aligned. At test time, a Mahalanobis positional score and a directional score computed against a bank of normal directions are combined multiplicatively, the directional score acting as a gate that amplifies the positional one where the local dynamics are atypical. Despite its simplicity, VACE achieves state-of-the-art performance on TSB-AD-M under rigorous evaluation, significantly outperforming more complex methods trained on substantially larger budgets.

版本更新 · v3arXiv:2602.23201

告诉我该学什么:让神经记忆可通过自然语言控制并具备泛化能力

Tell Me What To Learn: Generalizing Neural Memory to be Controllable in Natural Language

Max S. Bennett, Thomas P. Zollo, Richard Zemel

AI 速读 基于摘要

提出一种可用自然语言学习指令控制更新内容的广义神经记忆系统,使模型能从异构信息源中选择性学习。

cs.LG

查看原文摘要

Modern machine learning models are deployed in diverse, non-stationary environments where they must continually adapt to new tasks and evolving knowledge. Continual fine-tuning and in-context learning are costly and brittle, whereas neural memory methods promise lightweight updates with minimal forgetting. However, existing neural memory models typically assume a single fixed objective and homogeneous information streams, leaving users with no control over what the model remembers or ignores over time. To address this challenge, we propose a generalized neural memory system that performs flexible updates based on learning instructions specified in natural language. Our approach enables adaptive agents to learn selectively from heterogeneous information sources, supporting settings, such as healthcare and customer service, where fixed-objective memory updates are insufficient.

版本更新 · v2arXiv:2610.33301

The As-If Machine: Personalized AI counterfactuals make distant policy risks feel close

Joshua Ashkinaze, Yubo Zhou, Pengyu Chen, Avi Bhattacharya, Hyebin Park, Yanman Zhang, Stephanie Preston, Eric Gilbert 等 9 位作者

原文摘要

Life experiences can shape policy attitudes, but we only live one life. Citizens may therefore fail to support policies addressing risks they have never personally experienced. Narratives can make such risks immediate and personal, but personalized stories remain difficult to produce at scale. Here we ask whether AI can make distant scenarios feel concrete by simulating a participant's life as if that scenario occurred. We present the As-If Machine (AIM), a multi-agent system through which people co-write personalized counterfactual narratives---stories about their own lives under different policy futures. In two pre-registered and quota-matched experiments on science funding (N=1,007) and satellite cybersecurity (N=894), we compared AIM with two controls: reading an expert-written vignette on the same topic and co-writing with AIM about an unrelated topic. AIM increased psychological closeness---how close a risk feels to the self---by nearly identical margins in the two domains (0.36 and 0.38 standard deviations (SDs) against the vignette, 0.80 and 0.90 against unrelated writing). AIM increased psychological closeness even among participants with unfavorable views of AI. For policy attitudes, AIM's effects were 0.16 SDs against the vignette and 0.10 SDs (n.s.) against unrelated writing for science funding; the corresponding effects for satellite cybersecurity were 0.54 and 0.67 SDs. Across 23,496 human-AI autocomplete interactions, participants and the system played distinct roles: AIM supplied concrete detail that grounded the scenario, and participants supplied first-person perspective and emotion. Co-writing with the As-If Machine makes distant policy risks feel close.

cs.HC · cs.AI

版本更新 · v3arXiv:2511.16660

推理的认知基础及其在大语言模型中的表现

Cognitive Foundations for Reasoning and Their Manifestation in LLMs

Shuyue Stella Li, Priyanka Kargupta, Haocheng Wang, Jinu Lee, Dean Light, Shan Chen, Orevaoghene Ahia, Thomas L. Griffiths 等 12 位作者

AI 速读 基于摘要

该研究将认知科学整合为28个认知要素的分类体系,并据此细粒度比较18个模型与人类在推理过程中的认知要素使用差异。

cs.AI

查看原文摘要

Large language models (LLMs) solve complex problems yet fail on simpler variants, suggesting they achieve correct outputs through mechanisms fundamentally different from human reasoning. To understand this gap, we synthesize cognitive science research into a taxonomy of 28 cognitive elements spanning reasoning invariants, meta-cognitive controls, representations for organizing reasoning and knowledge, and transformation operations. We introduce a fine-grained evaluation framework and apply it to 192K traces from 18 models across text, vision, and audio, alongside 1,346 human think-aloud traces on a new, domain-agnostic dataset for human-model comparison. We find that models are able to deploy cognitive elements seen in human reasoning but differ in how they use them. Leveraging these patterns, we develop test-time reasoning guidance that automatically scaffolds successful reasoning topologies, improving the accuracy of the strongest models by 9.7 percentage points on ill-structured problems. Meta-analysis of 1.6K LLM reasoning papers reveals the research community concentrates on easily quantifiable elements but rarely targets self-awareness, whose use differs between human and model reasoning. By establishing a shared vocabulary between cognitive science and LLM research, our framework enables diagnosis of reasoning failures and principled development of models that reason through robust cognitive mechanisms rather than spurious shortcuts, while providing tools to test theories of human cognition at scale.

版本更新 · v2arXiv:2509.14956

Sentinel Agents for Secure and Trustworthy Agentic AI in Multi-Agent Systems

Diego Gosmar, Deborah A. Dahl

原文摘要

Multi-agent systems (MAS) built on large language models (LLMs) increasingly share a common conversational space, in which a single adversarial or erroneous message can propagate from one agent to many. This paper proposes Sentinel Agents, a dedicated security layer that monitors inter-agent communication with rule-based checks, behavioral analysis and LLM-based semantic inspection, and reports to a Coordinator Agent that enforces policy, for example by quarantining misbehaving agents. We describe the architecture, its deployment patterns and monitoring layers, and we evaluate a continuous-listener prototype on a balanced corpus of 1,000 synthetic messages (500 attacks across five families and 500 benign messages) with two local LLMs. LLM inspection with Gemma 4 12B reached F1 0.86 at a false-positive rate of 0.11, against 0.38 for rules alone. Combining components by logical OR did not improve on it, while a cascade in which only high-precision rules act on their own matched it. Calibrating the thresholds of the more conservative Qwen3 8B on a development split raised its held-out F1 from 0.56 to 0.80, and on multi-turn conversations the behavioral check detected patterns the rules missed but could not tell legitimate repetition from probing. The value of combining detectors therefore depends on the model, on its calibration and on the aggregation rule, and the results yield practical guidance for deploying Sentinel Agents.

cs.AI · cs.MA