2026-10论文中文摘要
更少监督，更好泛化：扩散编辑图像中的弱监督伪造区域定位
Less Supervision, Better Generalization: Weakly Supervised Fake Region Localization in Diffusion-Edited Images
arXiv 2609.36882 · 2026-10-02扩散策略改进：基于提案条件精炼流的方法
Diffusion Policy Improvement with Proposal-Conditioned Refinement Flows
arXiv 2609.36812 · 2026-10-02面向AI辅助工程的CAD原生Transformer算子
CANTO: CAD-Native Transformer Operators for AI-Aided Engineering
arXiv 2609.36806 · 2026-10-02NesTok：用于自回归图像生成的嵌套自对齐一维分词器
NesTok: Nested Self-Aligned 1D Tokenizer for Autoregressive Image Generation
arXiv 2609.36756 · 2026-10-02预测监督的方式决定VLA策略学习的内容
Where Predictive Supervision Goes Shapes What VLA Policies Learn
arXiv 2609.36645 · 2026-10-02什么使得高倍率知识可迁移？全切片成像中跨分辨率蒸馏的研究
What Makes High-Magnification Knowledge Transferable? A Study of Cross-Resolution Distillation in Whole-Slide Imaging
arXiv 2609.36407 · 2026-10-02由正一致性-不一致性配置刻画的分裂线的$n$-变量定理及其在保持问题中的应用
$n$-variable theorems for dividing lines characterized by positive consistency-inconsistency configurations, and their applications to preservation problems
arXiv 2609.36394 · 2026-10-02几乎线性极小序的典范性
On canonicity of almost linear minimal orders
arXiv 2609.36313 · 2026-10-02无名义词的证明：哥德尔本体论论证、其浅嵌入及《数学月刊》笔记中的开放问题
Proofs Without Nominals: Gödel's Ontological Argument, its Shallow Embedding, and the Open Questions of the Monatshefte Notes
arXiv 2609.36279 · 2026-10-02MERID：用于重度抑郁症分析的多模态递归自我改进智能体探索
MERID: Multimodal Exploration via Recursive Self-Improvement Agents for Major Depression Analysis
arXiv 2609.36235 · 2026-10-02Altermagnetism在La$_2$O$_3$Mn$_2$Se$_2$中产生暗激子和磁子的成对发射与吸收
Altermagnetism produces pair emission and absorption from dark excitons and magnons in La$_2$O$_3$Mn$_2$Se$_2$
arXiv 2609.36128 · 2026-10-02持续学习中的早期学习塑造表征变化的后续方向
Early Learning Shapes Later Directions Of Representation Change In Continual Learning
arXiv 2609.36081 · 2026-10-02Amadeus：当人的模型相遇
Amadeus: When Models of People Meet
arXiv 2609.35835 · 2026-10-02对齐预测：从训练数据预测错位
Alignment Forecasting: Predicting Misalignment From Training Data
arXiv 2609.35805 · 2026-10-02患者朝向对基于AI康复监测的单视角和多视角相机环境的影响
Impact of Patient Orientation in Single- and Multi-View Camera Environments for AI-based Rehabilitation Monitoring
arXiv 2609.35726 · 2026-10-02论确定性在多物品拍卖中的效力
On the Power of Determinism in Multi-Item Auctions
arXiv 2609.35711 · 2026-10-02跨越三千年的甲骨文演变追踪
Tracing the Evolution of Oracle Bone Characters Across Three Millennia
arXiv 2609.35674 · 2026-10-02基态制备的最优查询复杂度
Optimal Query Complexity for Ground-State Preparation
arXiv 2609.35668 · 2026-10-02关于 $p-1$ 和 $p+1$ 的最大素因子
On the largest prime factors of $p-1$ and $p+1$
arXiv 2609.35610 · 2026-10-02面向大语言模型的输出感知残差流剪枝
Output-aware Residual Stream Pruning for Large Language Models
arXiv 2609.35579 · 2026-10-02不要对所有内容进行接种：分层接种提示收窄后门触发器并保留期望特征
Don't Inoculate Everything: Stratified Inoculation Prompting Narrows Backdoor Triggers and Preserves Desired Traits
arXiv 2609.35356 · 2026-10-02莱顿人工智能与数学宣言
Leiden Declaration on Artificial Intelligence and Mathematics
arXiv 2609.35334 · 2026-10-02密度增强运动性和尺寸调节的群体感应
Quorum sensing with density-enhanced motility and size regulation
arXiv 2609.35085 · 2026-10-02Payne-Hanek 区间缩减算法的性能提升
A performance enhancement of the Payne-Hanek range reduction algorithm
arXiv 2609.35015 · 2026-10-02三角重采样用于长时程运动生成
Triangular Resampling for Long-Horizon Motion Generation
arXiv 2609.34697 · 2026-10-02奖励新颖演绎：求解器引导的过程奖励用于逻辑推理
Rewarding Novel Deductions: Solver-guided Process Supervision for Logical Reasoning
arXiv 2609.34660 · 2026-10-02无偏Top-$k$估计用于同策略蒸馏
Unbiased Top-$k$ Estimation for On-Policy Distillation
arXiv 2609.34447 · 2026-10-02相干性而非错误率主导多租户量子计算中的隐私
Coherence Rather Than Error Rate Governs Privacy in Multi-Tenant Quantum Computing
arXiv 2609.34411 · 2026-10-02TacGooseBumps (TacGB)：为仅法向触觉传感器加装剪切编码以学习接触丰富操作
TacGooseBumps (TacGB): Retrofitting Normal-Only Tactile Sensors with Shear Encoding for Learning Contact-Rich Manipulation
arXiv 2609.34006 · 2026-10-02视频，因此生成：通过时空类比统一视频任务
Unifying Video Tasks via Spatiotemporal Analogy
arXiv 2609.33935 · 2026-10-02Arnoux-Rauzy 词中的均匀因子平衡性
Uniform Factor-Balancedness in Arnoux-Rauzy Words
arXiv 2609.33919 · 2026-10-02少机器离线临时任务分配的EPTAS
An EPTAS for Vector Scheduling with Time Intervals
arXiv 2609.33493 · 2026-10-02学习销售：多产品市场中战略性大型语言模型智能体的强化学习
Learning to Sell: Reinforcement Learning for Strategic Large Language Model Agents in Multi-Product Markets
arXiv 2609.33289 · 2026-10-02FloodDiffusion 2：高效且路径可控的流式运动生成
FloodDiffusion 2: Efficient and Path Controllable Streaming Motion Generation
arXiv 2609.33167 · 2026-10-02技能实际做了什么？LLM智能体中工具与技能使用的估计目标与评估效度：一项批判性综述
What Does a Skill Actually Do? Estimands and Evaluation Validity for Tool and Skill Use in LLM Agents: A Critical Review
arXiv 2609.33153 · 2026-10-02不太难也不太易：从中间状态学习以进行LLM结构化推理
Not Too Hard, Not Too Easy: Learning from Intermediate States for LLM Structured Reasoning
arXiv 2609.33149 · 2026-10-02固定且仍不稳定：裁判内判决方差与LLM-as-Judge排行榜的噪声底限
Pinned and Still Unstable: Within-Judge Verdict Variance and the Noise Floor of LLM-as-Judge Leaderboards
arXiv 2609.33044 · 2026-10-02DEFINE：基于范例引导的口音控制用于零样本语音合成
DEFINE: Exemplar-Guided Accent Control for Zero-Shot TTS
arXiv 2609.32777 · 2026-10-02某些四阶非对称有限差分拉普拉斯算子的谱实性
Spectral Reality for Certain Fourth-Order Nonsymmetric Finite-Difference Laplacians
arXiv 2609.32410 · 2026-10-02带退火噪声水平的收敛即插即用图像恢复
Convergent Plug-and-Play Image Restoration with Annealed Noise Levels
arXiv 2609.32393 · 2026-10-02更少令牌，更多自我教学：面向极端视觉令牌缩减的在线策略自蒸馏
Fewer Tokens, More Self-Teaching: On-Policy Self-Distillation for Extreme Visual Token Reduction
arXiv 2609.32353 · 2026-10-02Devol-ONE：一个自回归Transformer混合体，统一视觉-语言-动作与潜在世界建模
Devol-ONE: One Autoregressive Mixture of Transformers to Unify Vision-Language-Action and Latent World Modeling
arXiv 2609.32193 · 2026-10-02DOHF：基于Doob的$h$-变换引导的在线扩散微调
DOHF: Online Diffusion Fine-tuning with Doob's $h$-transform Guidance
arXiv 2609.31882 · 2026-10-02用于GlueX-III粲夸克偶素计划的跃迁辐射探测器升级
Transition Radiation Detector Upgrade for the GlueX-III Charmonium Program
arXiv 2609.31585 · 2026-10-02面向复杂三维几何的谱/hp元方法的低阶细化预处理
Low-Order Refined Preconditioning for Spectral/hp Element Method for Complex, 3D Geometries
arXiv 2609.31443 · 2026-10-02精确求解子样本集成用于聚类回归：无需修剪水平的修剪
Group recovery after trimming, and level-free flagging, in robust clusterwise regression
arXiv 2609.31019 · 2026-10-02FRAM：轨迹引导的视觉特征选择用于紧凑型语言条件机器人操作
FRAM: Trajectory-Guided Visual Feature Selection for Compact Language-Conditioned Robot Manipulation
arXiv 2609.30965 · 2026-10-02Fermi-Pasta-Ulam-Tsingou系统的长时间Korteweg-de Vries近似
Long-time Korteweg-de Vries approximation for the Fermi-Pasta-Ulam-Tsingou system
arXiv 2609.30668 · 2026-10-02到/自自由模与内射模的不可约同态的结构
Structure of irreducible homomorphisms to/from free and injective modules
arXiv 2609.30638 · 2026-10-02旗流形的有理同伦论
Rational homotopy theory of flag manifolds
arXiv 2609.30591 · 2026-10-02
