2026-09论文中文摘要

• 深度ReLU表示的布朗头：激活质量与同样本选择的代价

  Brownian Heads for Deep ReLU Representations: Activation Mass and the Cost of Same-Sample Selection

  arXiv 2609.21422 · 2026-09-21

• 复标量场厚膜：线性扰动的稳定性及与引力耦合的标量Kaluza-Klein模式的演化

  Complex scalar field thick branes: stability of linear perturbation and evolution of scalar Kaluza-Klein modes coupled with gravity

  arXiv 2609.21421 · 2026-09-21

• 超自适应图与实现问题

  Super adaptable graphs and the realization problem

  arXiv 2609.21420 · 2026-09-21

• TrustBOM：一种跨组织的保密性SBOM可扩展架构

  TrustBOM: A Scalable Architecture for Confidentiality-Preserving SBOMs Across Organizations

  arXiv 2609.21419 · 2026-09-21

• 多能级弛豫下量子传感的极限信息率

  Ultimate Information Rate for Quantum Sensing under Multilevel Relaxation

  arXiv 2609.21418 · 2026-09-21

• 椭圆不动点处平面映射的规范形与首次积分

  Normal forms and first integrals of planar maps at elliptic fixed points

  arXiv 2609.21417 · 2026-09-21

• 面向性能优化的示教操作统一动态力引导框架

  A Unified Dynamic Force Guidance Framework for Performance-Optimized Kinesthetic Teaching

  arXiv 2609.21416 · 2026-09-21

• 洛伦兹模型中的局部细致平衡：以狂暴爆发取代蝴蝶效应

  Local Detailed Balance in the Lorenz Model: Replaces the Butterfly with Frenetic Bursting

  arXiv 2609.21415 · 2026-09-21

• 阿贝尔陈-西蒙斯理论中缺陷的函子理论

  A Functorial Theory of Defects in Abelian Chern-Simons Theory

  arXiv 2609.21414 · 2026-09-21

• 合成人类移动数据生成：表示、方法与实际能力的结构化综述

  Synthetic Human Mobility Data Generation: A Structured Review of Representations, Methods, and Practical Capabilities

  arXiv 2609.21413 · 2026-09-21

• 当在线适应有害时：参数冻结的测试时集成用于持续医学图像分割

  When Online Adaptation Hurts: Parameter-Frozen Test-Time Ensembling for Continual Medical Image Segmentation

  arXiv 2609.21412 · 2026-09-21

• 变脸耀变体B2 1420+326 (OQ 334)中的伽马射线准周期振荡

  Gamma-ray Quasi-periodic Oscillations in the Changing Look Blazar B2 1420+326 (OQ 334)

  arXiv 2609.21411 · 2026-09-21

• 预测胶结颗粒材料在化学损伤（脱粘）过程中的弹性质

  Predicting the Elastic Properties of a Cemented Granular Material during Chemical Damage (Debonding)

  arXiv 2609.21410 · 2026-09-21

• 地球内核中的异常物质势：暗物质与轻惰性中微子的一种现象学探针

  Anomalous Matter Potentials in Earth's Inner Core: A Phenomenological Probe of Dark Matter and Light Sterile Neutrinos

  arXiv 2609.21408 · 2026-09-21

• 量化感知卡尔曼估计用于扩散采样

  Quantization-Aware Kalman Estimation for Diffusion Sampling

  arXiv 2609.21407 · 2026-09-21

• 粒子与湍流之间的相对速度

  Relative velocity between a particle and turbulence

  arXiv 2609.21406 · 2026-09-21

• 利用动力系统极限环的强化学习优化流体混合

  Optimization of fluid mixing by reinforcement learning using limit cycles of a dynamical system

  arXiv 2609.21405 · 2026-09-21

• 基于SC-IMM教师信号的端到端自动驾驶轨迹输出稳定化

  Stabilizing Trajectory Outputs in End-to-End Autonomous Driving via SC-IMM Based Teacher Signals

  arXiv 2609.21404 · 2026-09-21

• SIRA：通过查询锚定对齐实现推理感知的手术器械分割

  SIRA: Reasoning-Aware Surgical Instrument Segmentation via Query-Anchored Alignment

  arXiv 2609.21402 · 2026-09-21

• 与机器交谈：人机对话中的道德、礼貌与对齐

  Talking Past the Machine: Morality, Politeness, and Alignment in Human-AI Dialogue

  arXiv 2609.21401 · 2026-09-21

• 场景语言模型用于开放词汇场景映射

  A Scene Language Model for Open-Vocabulary Scene Mapping

  arXiv 2609.21400 · 2026-09-21

• Riesz型不等式在粘合下的稳定性

  Stability of the Riesz-type inequalities under gluing

  arXiv 2609.21399 · 2026-09-21

• 部分纠缠熵的微分形式描述：协变相空间中Killing矢量构造的检验

  A Differential Form Description of Partial Entanglement Entropy: Testing a Killing Vector Construction in Covariant Phase Space

  arXiv 2609.21398 · 2026-09-21

• 与原始毕达哥拉斯四元组相关的 Frobenius 数

  Frobenius Numbers Associated with Primitive Pythagorean Quadruples

  arXiv 2609.21397 · 2026-09-21

• MarineCraft：通过模块化构建实现水下机器人的快速原型设计

  MarineCraft: Enabling Rapid Prototyping of Underwater Robots via Modular Construction

  arXiv 2609.21396 · 2026-09-21

• 常量列表插入-删除码：新界及对Levenshtein下界的改进

  Constant-List Insertion--Deletion Codes:New Bounds and an Improvement of Levenshtein's Lower Bound

  arXiv 2609.21395 · 2026-09-21

• 高重建质量与重启可重复性不保证恢复真实肌肉协同模式

  High Reconstruction Quality and Restart Repeatability Do Not Guarantee Recovery of Ground-Truth Muscle Synergies

  arXiv 2609.21394 · 2026-09-21

• 处理效应估计中的汇总指数

  Summary Indices in Treatment Effect Estimation

  arXiv 2609.21393 · 2026-09-21

• 全模态需求理解：多模态交互中上下文用户意图推断的基准

  Omni Demand Understanding: A Benchmark for Contextual User-Intent Inference in Multimodal Interaction

  arXiv 2609.21392 · 2026-09-21

• WS-NeRF：一种基于Mamba驱动的世界状态感知自适应去模糊神经辐射场

  WS-NeRF: A Mamba-Driven World-State-Aware Adaptive Deblurring Neural Radiance Field

  arXiv 2609.21391 · 2026-09-21

• 离线多模态大语言模型用于空中作战决策支持

  Offline Multimodal Large Language Models for Decision Support in Air Operations

  arXiv 2609.21390 · 2026-09-21

• 每个非完全图的有限删除诱导饱和性

  Finite deletion-induced saturation for every non-complete graph

  arXiv 2609.21388 · 2026-09-21

• 基于图方法的临床文档一致性再词汇化

  Consistent Relexicalization of Clinical Documents using Graph-Based Approach

  arXiv 2609.21387 · 2026-09-21

• AgentVidBench：用于评估MLLM智能体的多跳视频问答基准

  AgentVidBench: A Multi-Hop Video Question Answering Benchmark for Evaluating MLLM Agents

  arXiv 2609.21386 · 2026-09-21

• 关于从其他三角形全等准则重建SAS的研究，第二部分：消除驴桥定理

  On the Reconstruction of SAS from Other Triangle Congruence Criteria, Part II: Eliminating the Pons Asinorum

  arXiv 2609.21385 · 2026-09-21

• 对偶性、刚性及多段分解的剥离：关于Mitra、Offen与Sayag的猜想

  Duality, rigidity, and peeling for multisegments: on hypotheses of Mitra, Offen, and Sayag

  arXiv 2609.21384 · 2026-09-21

• 深度递归语言模型中的预测动力学

  Prediction Dynamics in Depth-Recurrent Language Models

  arXiv 2609.21383 · 2026-09-21

• 基于神经时间点过程的业务流程执行概率预测

  Probabilistic Forecasting of Business Process Executions with Neural Temporal Point Processes

  arXiv 2609.21382 · 2026-09-21

• 知识图谱增强的Chronos-2用于HEC-RAS代理预测

  Knowledge-Graph-Augmented Chronos-2 for HEC-RAS Surrogate Forecasting

  arXiv 2609.21381 · 2026-09-21

• 有限时长量子控制中的记忆信息窗口

  A memory-information window in finite-duration quantum control

  arXiv 2609.21380 · 2026-09-21

• JEPA引导扩散：用于生成式交通预测的预测性视觉-语言条件化

  JEPA Guided Diffusion: Predictive Vision-Language Conditioning for Generative Traffic Forecasting

  arXiv 2609.21379 · 2026-09-21

• ArenaFlow：从轨迹排序到开放式智能体强化学习的分层信用传播

  ArenaFlow: From Trajectory Ranking to Hierarchical Credit Propagation for Open-Ended Agent RL

  arXiv 2609.21378 · 2026-09-21

• AVT-Fabric：通过自适应证据选择实现主动视觉-触觉感知，用于高效机器人织物比较

  AVT-Fabric: Active Visuo-Tactile Perception via Adaptive Evidence Selection for Efficient Robotic Fabric Comparison

  arXiv 2609.21377 · 2026-09-21

• 拓扑光学频率梳

  Topological Optical Frequency Combs

  arXiv 2609.21375 · 2026-09-21

• 星系生态中重子循环的喷泉模式

  Fountain pattern of baryon cycle revealed in galaxy ecosystems

  arXiv 2609.21374 · 2026-09-21

• 光纤中偏振量子比特的无源泡利翻转：误差界与QKD性能

  Passive Pauli Toggling of Polarization Qubits in Optical Fibers: Error Bounds and QKD Performance

  arXiv 2609.21373 · 2026-09-21

• Cauchy交错定理对RLS算法收敛性的影响

  The Impact of the Cauchy Interlace Theorem on the Convergence of the RLS Algorithm

  arXiv 2609.21372 · 2026-09-21

• RobotEQ-Video：基于世界状态分类法的以视频为中心的社交主动智能基准

  RobotEQ-Video: A Video-Centric Benchmark for Social Proactive Intelligence with World-State Taxonomy

  arXiv 2609.21371 · 2026-09-21

• 具有一般边界条件的全息共形场论中的纠缠熵

  Entanglement entropy in holographic CFTs with generic boundaries

  arXiv 2609.21370 · 2026-09-21

• ProTracer：机器人操作中本体感觉引导的故障诊断

  ProTracer: Proprioception-Guided Failure Diagnosis in Robot Manipulation

  arXiv 2609.21369 · 2026-09-21