2026-10论文中文摘要
信号简化并非预测简化：诊断短期波动率中的残差神经预测
Signal Simplification Is Not Predictive Simplification: Diagnosing Residual Neural Forecasting in Short-Horizon Volatility
arXiv 2610.03019 · 2026-10-05裂变碎片的能量分辨自旋和宇称分布
Energy-Resolved Spin and Parity Distributions of Fission Fragments
arXiv 2610.03018 · 2026-10-05针对构音障碍和气管造口说话人的个性化自动语音识别：基于人工对话方法
Personalized Automatic Speech Recognition for a Dysarthric and Tracheostomic Speaker using Artificial Conversations
arXiv 2610.03017 · 2026-10-05从表达到反应：角色感知的视觉迁移与刺激引导的推理用于对话者情绪识别
From Expression to Reaction: Role-aware Visual Transfer and Stimulus-guided Reasoning for Interlocutor Emotion Recognition
arXiv 2610.03016 · 2026-10-05OmniAct3D：利用基础几何与证据接地推理实现全景3D检测
OmniAct3D: Leveraging Foundation Geometry and Evidence-Grounded Reasoning for Panoramic 3D Detection
arXiv 2610.03015 · 2026-10-05超越预定义汇点：面向LLM智能体的安全感知依赖分析
Beyond Predefined Sinks: Security-Aware Dependency Analysis for LLM Agents
arXiv 2610.03014 · 2026-10-05RYOPO：将端到端类别级物体姿态估计带入实时
RYOPO: Bringing End-to-End Category-Level Object Pose Estimation into Real Time
arXiv 2610.03013 · 2026-10-05重新思考固定时间网格：频率解耦的运动生成
Rethinking Fixed Temporal Grids: Frequency-Disentangled Motion Generation
arXiv 2610.03012 · 2026-10-05$L^p$ 框架下粘性守恒律系统的全局存在性
Global existence for viscous systems of conservation laws in $L^p$ framework
arXiv 2610.03011 · 2026-10-05工程可持续智能体：面向开发者工作流的智能体LLM系统比较研究
Engineering Sustainable Agents: A Systematic Comparison of Agentic LLMs for Developer Workflows
arXiv 2610.03010 · 2026-10-05非互易耦合诱导光机械系统中的非厄米纠缠
Nonreciprocal Coupling Induced Non-Hermitian Entanglement in Optomechanical Systems
arXiv 2610.03009 · 2026-10-05AvoKV-E：面向长推理的负载感知KV缓存驱逐
AvoKV-E: Payload-Aware KV Cache Eviction for Long Reasoning
arXiv 2610.03007 · 2026-10-05心理测量测试：量化低可靠性的后果并改进可靠性估计
Psychometric Tests: Quantifying the Consequences of Low Reliability and Improving Reliability Estimation
arXiv 2610.03006 · 2026-10-05Langevin型动力学与$\mathrm{SE}(3)$上的亚压缩性
Langevin-type dynamics and hypocoercivity on $\mathrm{SE}(3)$
arXiv 2610.03005 · 2026-10-05具有不同最大李雅普诺夫指数的若干开放台球系统
Some open billiards with distinct top Lyapunov exponents
arXiv 2610.03004 · 2026-10-05二元kagome超导候选材料中的拓扑电子态
Binary kagome superconducting candidates hosting topological electronic states
arXiv 2610.03003 · 2026-10-05统一多模态模型中的递归自我改进
Recursive Self-Improvement in Unified Multimodal Models
arXiv 2610.03002 · 2026-10-05神经数据需要语义标记化：行为事件作为会话可迁移标记的边界
Neural Data Needs Semantic Tokenization: Behavioral Events as Boundaries of Session-Transferable Tokens
arXiv 2610.03001 · 2026-10-05深度网络的时间几何：训练动力学的双曲表示用于内在可解释性
Temporal Geometry of Deep Networks: Hyperbolic Representations of Training Dynamics for Intrinsic Explainability
arXiv 2610.03000 · 2026-10-05方钴矿材料中增强和符号可逆的自旋霍尔与自旋能斯特电导率的从头设计
Ab initio design of enhanced and sign-reversible spin Hall and spin Nernst conductivity in skutterudite materials
arXiv 2610.02998 · 2026-10-05双qutrit Werner态总是局域的
Two-qutrit Werner state is always local
arXiv 2610.02997 · 2026-10-05面向可信人工智能的进化计算：从攻击与防御到自进化时代
Evolutionary Computation for Trustworthy AI: From Attacks and Defenses to Self-Evolving Era
arXiv 2610.02996 · 2026-10-05BISCEPTER：面向大规模系统软件的概率驱动二分查找
BISCEPTER: Probability-Driven Bisection for Large-Scale System Software
arXiv 2610.02995 · 2026-10-05Sentry：在测试时学习从LLM智能体故障中恢复
Sentry: Learning to Recover from LLM Agent Failures at Test Time
arXiv 2610.02994 · 2026-10-05与$Q$-曲率相关的几何流
A geometric flow associated to $Q$-curvature
arXiv 2610.02993 · 2026-10-05有界顶点连通度的正则图的极值谱隙
Extremal spectral gap of regular graphs with bounded vertex connectivity
arXiv 2610.02992 · 2026-10-05可微Koopman算子用于动态图上的对比学习
Differentiable Koopman Operator for Contrastive Learning on Dynamic Graphs
arXiv 2610.02990 · 2026-10-05系统级综合用于持续推进不确定性下快速机会约束故障后上升重规划
System Level Synthesis for Fast Chance-Constrained Post-Fault Ascent Replanning under Persistent Propulsion Uncertainty
arXiv 2610.02989 · 2026-10-05GRB 221009A 的尺寸演化与余辉模型的比较
Comparing the size evolution of GRB 221009A with afterglow models
arXiv 2610.02988 · 2026-10-05定量链扩张与整系数上同调填充
Quantitative chain extensions and integral cohomological fillings
arXiv 2610.02987 · 2026-10-05OLMo-Detect：一个用于大型语言模型成员推断的多阶段、混杂因素受控基准
OLMo-Detect: A Multi-Stage, Confounder-Controlled Benchmark for Membership Inference on Large Language Models
arXiv 2610.02986 · 2026-10-05具身智能体间的表征对齐
On Representational Alignment among Embodied Agents
arXiv 2610.02985 · 2026-10-05宇称-时间对称谐振子势阱中Kohn振荡的相互作用诱导周期偏移
Interaction-induced period shift of the Kohn oscillation in a parity-time-symmetric harmonic trap
arXiv 2610.02984 · 2026-10-05从连续到完全离散的Carleman估计：抛物格式的传递原理
From Continuous to Fully Discrete Carleman Estimates: A Transfer Principle for Parabolic Schemes
arXiv 2610.02983 · 2026-10-05PLCWorld：闭环工厂仿真中LLM生成的PLC程序的基准测试
PLCWorld: Benchmarking LLM-Generated PLC Programs in Closed-Loop Plant Simulation
arXiv 2610.02982 · 2026-10-05通过切割统计保障源自由域适应中的相互纠正
Safeguarding Mutual Correction in Source-Free Domain Adaptation via Cut Statistics
arXiv 2610.02981 · 2026-10-05爱因斯坦-朗之万源关联在封闭非相对论系统中的研究
Einstein--Langevin source correlations in closed non-relativistic systems
arXiv 2610.02980 · 2026-10-05RASPER：面向EHR结局预测的临床笔记奖励对齐摘要生成
RASPER: Reward-Aligned Summarization of Clinical Notes for EHR Outcome Prediction
arXiv 2610.02979 · 2026-10-05多带超导体太赫兹二维相干光谱中Leggett模式的边带指纹
Sideband fingerprint of the Leggett mode in terahertz two-dimensional coherent spectroscopy of multiband superconductors
arXiv 2610.02978 · 2026-10-05优化磁性纳米颗粒组装体中偶极相互作用的磁性热疗效率的结构依赖性
Optimizing the magnetic hyperthermia efficiency from the structural dependence of dipolar interactions in magnetic nanoparticles assemblies
arXiv 2610.02977 · 2026-10-05相关证据解码用于音频-视觉幻觉缓解
Relevant Evidence Decoding for Audio-Visual Hallucination Mitigation
arXiv 2610.02976 · 2026-10-05基于不完美代理奖励的可靠自进化
Reliable Self-Evolution with Imperfect Proxy Rewards
arXiv 2610.02975 · 2026-10-05从语言先验到领域自适应：用于可通行性估计的偏好学习
From Language Priors to Field Adaptation: Preference Learning for Traversability Estimation
arXiv 2610.02974 · 2026-10-05一种可应用的多元集路径序（扩展版）
An Applicative Multiset Path Order (Extended Version)
arXiv 2610.02973 · 2026-10-05CreateScore：基于领域理论的贝叶斯路由用于基于LLM的简历筛选
CreateScore: Domain-Theory-Informed Bayesian Routing for LLM-Based CV Screening
arXiv 2610.02972 · 2026-10-05追踪人类日常认知活动的脑电图与生物特征数据方法
Tracking Human Daily Cognitive Activity from EEG and Biometric Data
arXiv 2610.02971 · 2026-10-05基于规划器-评论家智能体修正机制的画像感知可信食谱生成
Profile-Aware Trustworthy Recipe Generation with Planner-Critic Agentic Remediation
arXiv 2610.02969 · 2026-10-05基于证据而非仅凭理由进行推理：面向基于LLM的推荐的可验证偏好证明
Reasoning with Evidence, Not Merely Rationales: Verifiable Preference Proofs for LLM-Based Recommendation
arXiv 2610.02968 · 2026-10-05通过组合偏好与规则奖励对前沿文生图模型进行后训练
Post-Training Frontier Text-to-Image Models by Composing Preference and Rubric Rewards
arXiv 2610.02967 · 2026-10-05XMM-Newton解析PeVatron微类星体V4641 Sgr中的秒差距尺度X射线喷流
XMM-Newton Resolves Parsec-scale X-ray Jets in the PeVatron Microquasar V4641 Sgr
arXiv 2610.02966 · 2026-10-05
