2026-09论文中文摘要
从机器人足-地力历史推断土壤内摩擦角：一种基于贝叶斯反演的本体感觉土壤传感方法
Inferring Soil Friction Angle from Robot Foot-Ground Force Histories: A Bayesian Inverse Approach to Proprioceptive Soil Sensing
arXiv 2609.36582 · 2026-09-30MemEvo：流式视频记忆机制的自动发现
MemEvo: Automatic Discovery of Streaming Video Memory Mechanisms
arXiv 2609.36581 · 2026-09-30被动高斯计量学中松弛性的根系结构
Root-system structure of sloppiness in passive Gaussian metrology
arXiv 2609.36579 · 2026-09-30因子化调度原理：通过结构化加性函数学习可解释且可迁移的策略
Factorized Scheduling Principle: Learning Interpretable and Transferable Policies via Structured Additive Functions
arXiv 2609.36578 · 2026-09-30音频语言模型中基于长期记忆引导的目标感知增强
Long-Term Memory-Guided Enhancement for Target Perception in Audio-Language Models
arXiv 2609.36577 · 2026-09-30分而注入：智能体能否从片段中重构间接提示注入？
Divide and Inject: Can Agents Reconstruct an Indirect Prompt Injection from Fragments?
arXiv 2609.36576 · 2026-09-30EquivDP3：用于数据高效人形机器人移动操作任务的SIM(3)不变点云编码器
EquivDP3: A SIM(3)-Invariant Point-Cloud Encoder for Data-Efficient Humanoid Loco-Manipulation
arXiv 2609.36575 · 2026-09-30共享相位算术用于并行量子旋转
Shared Phase Arithmetic for Parallel Quantum Rotations
arXiv 2609.36574 · 2026-09-30CyberPersistBench：评估基于LLM的网络攻击者的安装与持久化能力
CyberPersistBench: Evaluating LLM-Based Cyber Attackers on Installation and Persistence
arXiv 2609.36573 · 2026-09-30量子Fisher信息与量子相干性之间的因子化动力学
Factorization dynamics between quantum Fisher information and quantum coherence
arXiv 2609.36571 · 2026-09-30CounterSteer：通过激活引导抑制间接提示注入
CounterSteer: Suppressing Indirect Prompt Injection with Activation Steering
arXiv 2609.36570 · 2026-09-30从检查点变化到监督微调中的选择增益
From Checkpoint Variation to Selection Gains in Supervised Fine-Tuning
arXiv 2609.36569 · 2026-09-30非负Ricci曲率下黎曼扩散的尖锐收敛性与采样权衡
Sharp Convergence and Sampling Trade-offs for Riemannian Diffusion under Nonnegative Ricci Curvature
arXiv 2609.36568 · 2026-09-30Boltzmann 关系下 Navier--Stokes--Fourier--Poisson 系统的激波剖面
Shock profiles for the Navier--Stokes--Fourier--Poisson system under the Boltzmann relation
arXiv 2609.36567 · 2026-09-30关于Walker卫星星座路由表缺乏周期性的研究
On the Lack of Periodicity of Walker Satellite Constellation Routing Tables
arXiv 2609.36566 · 2026-09-30迈向生成式视频通信：一种双流数字传输框架
Toward Generative Video Communication: A Dual-Stream Digital Transmission Framework
arXiv 2609.36565 · 2026-09-30平面带状等周问题的比较定理
A comparison theorem for the planar strip isoperimetric problem
arXiv 2609.36564 · 2026-09-30AffectReveal：超越视觉表象的事件锚定情感识别
AffectReveal: Event-Grounded Emotion Recognition Beyond Visual Appearances
arXiv 2609.36563 · 2026-09-30ThinkingGuard：在多模态大语言模型中通过逐步风险归因解码隐含危害
ThinkingGuard: Decoding Implicit Hazards via Step-by-Step Risk Attribution in Multimodal Large Language Models
arXiv 2609.36562 · 2026-09-30黎曼样条回归模型：用于在任意黎曼流形上学习时间场
Riemannian Splat Regression Models for Learning Time Fields on Arbitrary Riemannian Manifolds
arXiv 2609.36561 · 2026-09-30FM-ReID：面向目标重识别的选择性竞争令牌路由
FM-ReID: Selective Competitive Token Routing for Object Re-Identification
arXiv 2609.36560 · 2026-09-30HiTS-CL：面向长时域知识图谱外推的持续学习框架
HiTS-CL: A Continual Learning Framework for Long-Horizon Temporal Knowledge Graph Extrapolation
arXiv 2609.36559 · 2026-09-30一种用于同时检测压力和粘附性的鲁棒单传感元件触觉传感器，具有实时信号解耦能力
A robust single-sensing-element tactile sensor for concurrent pressure and tackiness detection with real-time signal decoupling capability
arXiv 2609.36558 · 2026-09-30医学视觉语言模型如何未能充分利用其视觉编码器：皮肤病学视角
How Medical VLMs Underutilize Their Vision Encoders: A Dermatology Perspective
arXiv 2609.36557 · 2026-09-30MAADBench：多智能体系统异常检测的可刷新范式
MAADBench: The Refreshable Paradigm for Anomaly Detection in Multi-Agent Systems
arXiv 2609.36556 · 2026-09-30广义有效自旋链形式体系：一维光晶格中的多组分任意子
Generalized Effective Spin-Chain formalism for multicomponent anyons in one-dimensional optical lattices
arXiv 2609.36555 · 2026-09-30一维光晶格中的多分量任意子
Multicomponent anyons in one-dimensional optical lattices
arXiv 2609.36554 · 2026-09-30学习探索铰接物体操作中的隐藏运动学
Learning to Explore Hidden Kinematics for Articulated Object Manipulation
arXiv 2609.36553 · 2026-09-30SERA：用于最大似然强化学习的尺度均衡回放分配
SERA: Scale-Equalized Rollout Allocation for Maximum Likelihood Reinforcement Learning
arXiv 2609.36552 · 2026-09-30有向图的 $k$-强定向的度条件
Degree conditions for $k$-strong orientations of digraphs
arXiv 2609.36551 · 2026-09-30基于检索的修订与先验注入：探测检索增强的专利权利要求修改
Grounded Revision vs. Prior Injection: Probing Retrieval-Augmented Patent Claim Amendment
arXiv 2609.36550 · 2026-09-30MedForge-RSI：通过递归自我改进实现医学深度伪造检测
MedForge-RSI: Medical Deepfake Detection via Recursive Self-Improvement
arXiv 2609.36549 · 2026-09-30高阶形式BRST反常的经典阴影
Classical Shadows of Higher-Form BRST Anomalies
arXiv 2609.36548 · 2026-09-30单体素无线NeRF用于空间频谱预测
Single-Voxel Wireless NeRF for Spatial Spectrum Prediction
arXiv 2609.36547 · 2026-09-30交互式策略蒸馏与双向提出-验证机制
Interactive-Policy Distillation with Bidirectional Propose-and-Verify
arXiv 2609.36546 · 2026-09-30DraftTrace：面向AI辅助写作的多视图分析环境
DraftTrace: A Multi-View Analytics Environment for AI-Integrated Writing
arXiv 2609.36544 · 2026-09-30面向规划的3D场景补全：基于部分观测的耦合TUDF占用表示学习
Planning Oriented 3D Scene Completion via Coupled TUDF Occupancy Representation Learning from Partial Observations
arXiv 2609.36543 · 2026-09-30系综平均中的涌现对称性
Emergent Symmetries in Ensemble Averages
arXiv 2609.36542 · 2026-09-30量子参考相对于经典参考在泄漏检测中的指数优势
Exponential Advantage of Quantum over Classical References in Leakage Detection
arXiv 2609.36541 · 2026-09-30反应式实时流策略：通过异步分布对齐
Reactive Real-Time Flow Policies via Asynchronous Distribution Alignment
arXiv 2609.36540 · 2026-09-30连续时间系统的组合不变性：从光滑性到Lebesgue密度
Invariance is Compositional for Continuous-time Systems: From Sleekness to Lebesgue Density
arXiv 2609.36539 · 2026-09-30受挫 Gd3+ 双钙钛矿作为亚 100 mK 绝热退磁制冷的高性能磁热材料
Frustrated Gd3+ Double Perovskites as High-Performance Magnetocaloric Materials for Sub-100 mK Adiabatic Demagnetization Refrigeration
arXiv 2609.36538 · 2026-09-30不对称侦察员-工人侦察用于未知环境中的路线验证
Asymmetric Scout-Worker Reconnaissance for Route Validation in Unknown Environments
arXiv 2609.36537 · 2026-09-30传感器丢失下基于扩散先验的KSTAR平衡重建
Diffusion prior for KSTAR equilibrium reconstruction under sensor dropout
arXiv 2609.36536 · 2026-09-30当更新不再是学习：通过可学习信息增益重新思考LLM自我进化
When Updating Stops Being Learning: Rethinking LLM Self-Evolution via learnable information gain
arXiv 2609.36535 · 2026-09-30查询中身份信号对检索的敏感性
Retrieval Sensitivity to Identity Signals in Queries
arXiv 2609.36534 · 2026-09-30基于超导加速器的脉冲慢正电子束时间结构计算
Calculation of the temporal structure for a pulsed slow positron beam based on superconducting accelerator
arXiv 2609.36533 · 2026-09-30分层效用校准用于结构化多类决策
Hierarchical Utility Calibration for Structured Multiclass Decisions
arXiv 2609.36532 · 2026-09-30轨迹级模态引导的可控扩散多机器人运动规划
Trajectory-Level Mode Guidance for Controllable Diffusion-Based Multi-Robot Motion Planning
arXiv 2609.36530 · 2026-09-30三元线性注意力：用于长上下文序列建模的三维循环状态
Triadic Linear Attention: Three-Dimensional Recurrent States for Long-Context Sequence Modeling
arXiv 2609.36529 · 2026-09-30
