2026-08论文中文摘要
智能体工作流的全局优化与推理时区域嫁接
Global Optimization and Inference-Time Region Grafting for Agentic Workflows
arXiv 2608.02353 · 2026-08-04Qwen-CUA：面向几乎所有场景的原生计算机使用智能体
Qwen-CUA: Native Computer Use for (almost) Everything
arXiv 2608.02352 · 2026-08-04KC-Agent：用于高效机器学习模型改进的双进程认知架构
KC-Agent: A Dual-Process Cognitive Architecture for Efficient ML Model Improvement
arXiv 2608.02351 · 2026-08-04用于二进制程序聚类的自监督表示：从实证研究到检索增强学习
Self-Supervised Representations for Binary Program Clustering: From Empirical Study to Retrieval-Augmented Learning
arXiv 2608.02348 · 2026-08-04低温聚焦离子束微结构加工技术实现Tl₂Ba₂CuO₆₊δ（Tl2201）的c轴输运定量测量
Cryogenic focused-ion-beam microstructuring enabling quantitative $c$-axis transport measurements in Tl$_2$Ba$_2$CuO$_{6+δ}$
arXiv 2608.02344 · 2026-08-04硬约束、平滑梯度：通过可微投影学习可行库存策略
Hard Constraints, Smooth Gradients: Learning Feasible Inventory Policies via Differentiable Projection
arXiv 2608.02343 · 2026-08-04拓扑材料NbNiTe5中压力诱导的非晶化与超导性共存
Pressure-induced concurrent amorphization and superconductivity in topological material NbNiTe5
arXiv 2608.02342 · 2026-08-04Wi-Fi中的广播速率限制：协作式边缘大语言模型（LLM）推理被遗忘的瓶颈
Broadcast Rate Limits in Wi-Fi: A Forgotten Bottleneck for Collaborative Edge LLM Inference
arXiv 2608.02341 · 2026-08-04非局部算子的逐点端点极限
Pointwise endpoint limits for nonlocal operators
arXiv 2608.02340 · 2026-08-04奇诱导子图与奇偶诱导子图的界
Bounds on Odd and Odd-Even Induced Subgraphs
arXiv 2608.02339 · 2026-08-04亚线性时间内恢复植入着色
Recovering Planted Colorings in Sublinear Time
arXiv 2608.02338 · 2026-08-04少层CrSBr中磁序的光控切换
Optical switching of magnetic order in few-layer CrSBr
arXiv 2608.02337 · 2026-08-04在权限中迷失：探索Microsoft 365应用生态系统
Lost in Permissions: Exploring the Microsoft 365 App Ecosystem
arXiv 2608.02336 · 2026-08-04二分图的DP着色函数
The DP Color Function of Bipartite Graphs
arXiv 2608.02335 · 2026-08-04面向离线强化学习的带行为优势修正的扩散策略
Diffusion Policy with Behavioral Advantage Correction for Offline Reinforcement Learning
arXiv 2608.02332 · 2026-08-04用于野外身体情绪表达的上下文感知领域专家混合模型
Context-Aware Mixture of Domain Experts for Bodily Expression of Emotion in the Wild
arXiv 2608.02331 · 2026-08-04超临界区域中MHD系统的强不适定性：无粘性与粘性情形
Strong ill-posedness for the MHD system in the supercritical regime: inviscid and viscous
arXiv 2608.02330 · 2026-08-04人工智能生成代码的自动化且可解释的来源研究
On Automated and Explainable Provenance of AI-Generated Code
arXiv 2608.02329 · 2026-08-04用于纳米光子学和量子技术的绝缘体上磷化镓
Gallium phosphide on insulator for nanophotonics and quantum technologies
arXiv 2608.02328 · 2026-08-04可学习但不可模拟：学习模型的量子资源理论
Learnable yet not simulable: a quantum resource theory of learning models
arXiv 2608.02325 · 2026-08-04用于多模态纵向图像插补与插值的隐式神经表示
Implicit Neural Representations for Multimodal Longitudinal Image Imputation and Interpolation
arXiv 2608.02324 · 2026-08-04面向NDVI时间序列重建的全球尺度自监督时空学习
Global-Scale Self-Supervised Spatiotemporal Learning for NDVI Time-Series Reconstruction
arXiv 2608.02322 · 2026-08-04FastGFDs：基于Desbordante的高效图函数依赖验证
FastGFDs: Efficient Validation of Graph Functional Dependencies with Desbordante
arXiv 2608.02321 · 2026-08-04TravKAN：基于Kolmogorov-Arnold网络的快速可解释非线性可通行性分析
TravKAN: Fast and Interpretable Nonlinear Traversability Analysis with Kolmogorov-Arnold Networks
arXiv 2608.02320 · 2026-08-04多尺度随机环境中常微分方程的平均原理
Averaging Principle for Ordinary Differential Equations in a Multiscale Random Environment
arXiv 2608.02319 · 2026-08-04带噪声的k-means++并非过于嘈杂
Noisy k-means++ is Not too Noisy
arXiv 2608.02318 · 2026-08-04双调和柯西问题的最小二乘弱Galerkin方法
A Least-Squares Weak Galerkin Method for the Biharmonic Cauchy Problem
arXiv 2608.02317 · 2026-08-04基于残差的自适应卡尔曼滤波用于四足机器人状态估计
Residual-Based Adaptive Kalman Filtering for Legged Robot State Estimation
arXiv 2608.02316 · 2026-08-04GEOID-Flood：用于洪水分割的大规模多模态基准数据集
GEOID-Flood: A Large-Scale Multi-Modal Benchmark Dataset for Flood Segmentation
arXiv 2608.02315 · 2026-08-04胶体系统
Colloidal Systems
arXiv 2608.02313 · 2026-08-04局部环上${}^2 D_\ell$（$\ell \geq 4$）型扭Chevalley群的自同构
Automorphisms of Twisted Chevalley Groups of type ${}^2 D_\ell \ (\ell \geq 4)$ over Local Rings
arXiv 2608.02312 · 2026-08-04用于健康虚假信息验证的基于证据的检索增强Transformer框架
An Evidence-Grounded Retrieval-Augmented Transformer Framework for Health Misinformation Verification
arXiv 2608.02310 · 2026-08-04CalibBEV：基于鸟瞰图对齐的激光雷达-相机标定方法
CalibBEV: LiDAR-Camera Calibration via BEV Alignment
arXiv 2608.02309 · 2026-08-04基于集合的通用认知状态估计框架：理论与在有条件自动驾驶中的应用
A General Set-Based Framework for Cognitive State Estimation: Theory and Application to Conditionally Automated Driving
arXiv 2608.02308 · 2026-08-04扩散-弛豫系统中的平衡化与局域化
Equilibration versus Localization in a Diffusion-Relaxation System
arXiv 2608.02307 · 2026-08-04用于动态形状连续且鲁棒比较的前推变换
The Push-Forward Transform for Continuous and Robust Comparison of Dynamic Shapes
arXiv 2608.02306 · 2026-08-04BRiG-AFA: 面向非近视主动特征获取的Bellman风险-去向学习
BRiG-AFA: Bellman Risk-to-Go Learning for Non-Myopic Active Feature Acquisition
arXiv 2608.02305 · 2026-08-04thagomizer拟阵与K_{2,n}图拟阵的Kazhdan--Lusztig多项式及Z-多项式的实根性
Real-rootedness of Kazhdan--Lusztig and $Z$-polynomials of thagomizer matroids and graphic matroids of $K_{2,n}$
arXiv 2608.02303 · 2026-08-04自分割轨迹：智能体声明边界作为训练单元
Trajectories That Segment Themselves: Agent-Declared Boundaries as a Training Unit
arXiv 2608.02302 · 2026-08-04平衡态千兆赫声学揭示液体中的长程限域效应
Equilibrium gigahertz acoustics reveals long-range confinement in liquids
arXiv 2608.02301 · 2026-08-04通用视觉语言模型（VLM）可指导天文基础模型更好地识别星系形态
A General-Purpose VLM Can Teach an Astronomy Foundation Model to Better Recognize Galaxy Morphology
arXiv 2608.02300 · 2026-08-04有序图的拉姆齐重数
Ramsey multiplicity for ordered graphs
arXiv 2608.02299 · 2026-08-04金属的长短时线性响应：一种几何方法
Long and short time linear response of metals: a geometric approach
arXiv 2608.02298 · 2026-08-04带记忆的量子重置
Quantum resetting with memory
arXiv 2608.02297 · 2026-08-04TrainShield：面向网络安全培训的目标感知机制
TrainShield: Targeted Awareness for Cybersecurity Training
arXiv 2608.02296 · 2026-08-04MechGeo：在Lean 4中自动形式化与证明欧几里得几何
MechGeo: Autoformalizing and Proving Euclidean Geometry in Lean 4
arXiv 2608.02295 · 2026-08-04XRISM 解析 NGC 4945 的核周环境
XRISM Resolves the Circum-nuclear Environment of NGC 4945
arXiv 2608.02294 · 2026-08-04创作者经济平台上实现收益最大化的动态流量分配
Dynamic Traffic Allocation for Revenue Maximization on Creator Economy Platform
arXiv 2608.02293 · 2026-08-04共享前缀，更优信用：面向多智能体推理的自适应路由
Shared Prefixes, Better Credit: Adaptive Routing for Multi-Agent Reasoning
arXiv 2608.02291 · 2026-08-04SpikeRestormer：通过统一事件推理实现节能的全合一图像复原
SpikeRestormer: Towards Energy-Efficient All-in-One Image Restoration via Unified Event Reasoning
arXiv 2608.02290 · 2026-08-04
