2026-10论文中文摘要
双彗星猜想的一个有限阈值
A finite threshold for the double-comet conjecture
arXiv 2610.02913 · 2026-10-05$\zeta(2)$ 的无理指数的一个新上界
A new upper bound for the irrationality exponent of $ζ(2)$
arXiv 2610.02912 · 2026-10-05探测训练框架：语言模型陈旧数据强化学习比较中的设置检查
Probe the Harness: Setup Checks for Stale-Data RL Comparisons in Language Models
arXiv 2610.02911 · 2026-10-05频率不等于敏感性：识别稀疏MoE大语言模型中的安全敏感专家
Frequency Is Not Sensitivity Identifying Safety-Sensitive Experts in Sparse MoE LLM
arXiv 2610.02910 · 2026-10-05约束感知训练
Constraint-Aware Training
arXiv 2610.02909 · 2026-10-05排列不变的多重叠干扰信号时频分割与参数估计
Permutation-Invariant Time-Frequency Segmentation and Parameter Estimation of Multiple Overlapping Interference Signals
arXiv 2610.02908 · 2026-10-05残差网络是否路由？残差网络中的稀疏交互专家
Do ResNets Route? Sparse Interaction Experts in Residual Networks
arXiv 2610.02907 · 2026-10-05切线 Schrödinger 桥匹配：利用机制敏感性学习随机输运
Tangent Schrödinger Bridge Matching: Learning Stochastic Transport with Mechanistic Sensitivities
arXiv 2610.02906 · 2026-10-05制备顺序控制钨上硼与锂之间的氧分配
Preparation sequence controls oxygen partitioning between boron and lithium on tungsten
arXiv 2610.02905 · 2026-10-05分散式稳定性证书的系统性SDP搜索
Systematic SDP Search for Decentralized Stability Certificates
arXiv 2610.02904 · 2026-10-05ViTok：在AM-RADIO风格多教师蒸馏中利用PHI-S和掩码图像建模提升密集语义
ViTok: Improving Dense Semantics in AM-RADIO-Style Multi-Teacher Distillation with PHI-S and Masked Image Modelling
arXiv 2610.02903 · 2026-10-05共识优化算法近期变体的定量平均场极限
Quantitative Mean-Field Limits for Recent Variants of Consensus-Based Optimization Optimization
arXiv 2610.02901 · 2026-10-05均匀强偏折透镜效应在近极端克尔黑洞中的应用
Uniform Strong-Deflection Lensing in Near-Extremal Kerr
arXiv 2610.02900 · 2026-10-05正则边界条件与解析K-同调
Regular boundary conditions and analytic K-homology
arXiv 2610.02899 · 2026-10-05MixVLA：面向泛化视觉-语言-动作模型的非不变信息自适应混合
MixVLA: Adaptive Mixing of Non-Invariant Information for Generalizable Vision-Language-Action Models
arXiv 2610.02898 · 2026-10-05写入时解释：多目标智能体记忆的策略消融
Interpreting at Write Time: A Policy Ablation for Multi-Goal Agent Memory
arXiv 2610.02897 · 2026-10-05辐射计效应对无限薄圆盘的作用
Radiometer effect on an infinitely thin circular disk
arXiv 2610.02896 · 2026-10-05非马尔可夫纠缠通过参数介导的分子腔光力学实现
Non-Markovian Entanglement via Parametrically Mediated Molecular Cavity Optomechanics
arXiv 2610.02895 · 2026-10-05何时可以信任匹配原则？有限样本和模型不确定性下的稳健部署几何
When Can We Trust the Matching Principle? Robust Deployment Geometry Under Finite-Sample and Model Uncertainty
arXiv 2610.02894 · 2026-10-05高维数据的残差树高斯过程建模框架
A Residual Tree Gaussian Process Modeling Framework for High-Dimensional Data
arXiv 2610.02893 · 2026-10-05Rank-One HCIZ 积分与乘法卷积
Rank-One HCIZ Integrals and Multiplicative Convolutions
arXiv 2610.02892 · 2026-10-05高效且精确的升阶-解耦广义可迁移神经网络用于椭圆和移动界面问题
Efficient and Accurate Lift-and-Decoupling Generalized Transferable Neural Networks for Elliptic and Moving Interface Problems
arXiv 2610.02891 · 2026-10-05基于滑模控制的模型化扰动抑制：连续时间公式与恰当隐式离散化
Model-Based Disturbance Rejection via Sliding-Mode Control: Continuous-Time Formulation and Proper Implicit Discretization
arXiv 2610.02890 · 2026-10-05超长伽马射线暴GRB 220627A的研究：探索可能的蓝超巨星坍缩星起源
Study of the Ultra-Long GRB 220627A: Exploring a Possible Blue Supergiant Collapsar Origin
arXiv 2610.02889 · 2026-10-05复杂网络上的拓扑绝缘体
Topological insulators on complex networks
arXiv 2610.02888 · 2026-10-05揭示多模态大语言模型中的认知不确定性：基于因果不变掩码的方法
Revealing Epistemic Uncertainty in MLLMs via Causal-Invariant Masking
arXiv 2610.02887 · 2026-10-05无触发器的错误信息：从事实性回答到下游决策
Misinformation Without Triggers: From Factual Answers to Downstream Decisions
arXiv 2610.02886 · 2026-10-05PsyEvo：一种在测试时自我进化的个性化心理咨询代理
PsyEvo: A Personalized Counseling Agent That Self-Evolves at Test Time
arXiv 2610.02885 · 2026-10-05未缠结聚合物熔体中非线性剪切流变的取向起源
Orientational origin of nonlinear shear rheology in unentangled polymer melts
arXiv 2610.02884 · 2026-10-05复杂网络的结构交叉：桥接度相关性与分形性
Structural crossover of complex networks: bridging degree correlation and fractality
arXiv 2610.02883 · 2026-10-05DyRA：用于深度神经网络中高效矩阵乘法的动态残差逼近
DyRA: Dynamic Residual Approximation for Efficient Matrix Multiplication in DNNs
arXiv 2610.02882 · 2026-10-05迈向全模态图基础模型：一种拓扑驱动的绑定方法
Toward Omni Multimodal Graph Foundation Model: A Topology-Driven Binding Approach
arXiv 2610.02881 · 2026-10-05已找到但未读取：当提取文本弥合文档视觉语言模型中的检索-读取差距时
Found but Not Read: When Extracted Text Closes the Retrieval-Reading Gap in Document Vision-Language Models
arXiv 2610.02880 · 2026-10-05相对论惯性系中的PaxosLease
PaxosLease in Relativistic Inertial Frames
arXiv 2610.02879 · 2026-10-05基于合作原则评估视觉语言模型中的视觉问答
Evaluating VQA in Vision Language Models using Cooperative Principles
arXiv 2610.02878 · 2026-10-05评估超越分数对齐的LLM作为评判者：残差评判难度的心理测量学分析
Evaluating LLM-as-a-Judge Beyond Score Alignment: A Psychometric Analysis of Residual Judging Difficulty
arXiv 2610.02877 · 2026-10-05看见、说出但不使用：多模态大语言模型中从可报告的空间事实到可用状态
Seeing, Saying, but Not Using: From Reportable Spatial Facts to Usable States in Multimodal Large Language Models
arXiv 2610.02876 · 2026-10-05查询感知路由：编码器中跨语言性能的提升
Query-aware routing for Cross-lingual performance gains in Encoders
arXiv 2610.02875 · 2026-10-05SceneFactory-3D：将2D交通场景提升为3D物理反事实，用于可扩展的物理基础安全评估
SceneFactory-3D: Lifting 2D Traffic Scenes into 3D Physical Counterfactuals for Scalable Physically Grounded Safety Evaluation
arXiv 2610.02874 · 2026-10-05ConvoDrift：用于建模文体语气演变的多轮对话数据集
ConvoDrift: A Multi-Turn Conversational Dataset for Modeling Stylistic Tone Evolution
arXiv 2610.02873 · 2026-10-05符号网络中的同伴效应：区分正负关系的影响
Peer Effects in Signed Networks: Separating Influence Through Positive and Negative Ties
arXiv 2610.02872 · 2026-10-05利用斑点强度轮廓对纳米束旋进电子衍射图案进行取向精化
Orientation refinement for nanobeam precession electron diffraction patterns using spot intensity profiles
arXiv 2610.02871 · 2026-10-05Azumaya下降与étale局部循环同调比较：一个非交换滤过
Azumaya descent and étale-local cyclotomic comparison: a noncommutative filtration
arXiv 2610.02870 · 2026-10-05AgentTrap：针对自主渗透测试代理的状态性反馈欺骗
AgentTrap: Stateful Feedback Deception against Autonomous Penetration Testing Agents
arXiv 2610.02869 · 2026-10-05子群体偏移与异常值污染下的分布鲁棒生存模型
Distributionally Robust Survival Models under Subpopulation Shift and Outlier Contamination
arXiv 2610.02868 · 2026-10-05TACD：通过终端放大控制蒸馏高效文本到运动模型
TACD: Distilling Efficient Text-to-Motion Models via Terminal Amplification Control
arXiv 2610.02867 · 2026-10-05DIVINE：通过多样指标重建实现简单的跨市场股票预训练
DIVINE: Simple Cross-Market Stock Pretraining via Diverse Indicator Reconstruction
arXiv 2610.02866 · 2026-10-05关于时间序列预测中的机器遗忘研究
On Unlearning for Time-series Forecasting
arXiv 2610.02865 · 2026-10-05NeuroLens：从慢性记录中学习神经语义的潜在嵌入
NeuroLens: Learning Latent Embeddings of Neural Semantics from Chronic Recordings
arXiv 2610.02864 · 2026-10-05多智能体人工智能作为私人财富管理中的嵌套委托-代理问题：瑞士、德国和奥地利的授权代表与证据控制
Multi-Agent AI as a Nested Principal-Agent Problem in Private Wealth Management: Mandate Representation and Evidence Control in Switzerland, Germany and Austria
arXiv 2610.02863 · 2026-10-05
