2026-08论文中文摘要
磁化与非磁化等离子体中电子-离子碰撞输运的第一性原理模拟
First-Principles Simulation of Electron-Ion Collisional Transport in Magnetized and Unmagnetized Plasmas
arXiv 2608.24012 · 2026-08-26SAGE：面向中国古代文献理解的从直接回答到证据导向推理
SAGE: From Direct Answering to Evidence-Grounded Inference for Chinese Ancient Document Understanding
arXiv 2608.24011 · 2026-08-26吸收梯度冲突：通过 Kent 分布建模语义方差用于跨模态哈希
Absorbing Gradient Conflicts: Modeling Semantic Variance via Kent Distributions for Cross-Modal Hashing
arXiv 2608.24010 · 2026-08-26微引力透镜图像被透镜周盘掩蔽：对巡天的启示
Occultation of microlensed images by circum-lens disks: Implications for surveys
arXiv 2608.24009 · 2026-08-26线性模式区域下的阈值与奇偶玻色采样
Threshold and Parity BosonSampling in the Linear-Mode Regime
arXiv 2608.24008 · 2026-08-26狄拉克界面态的工程化
Engineering Dirac interface states
arXiv 2608.24006 · 2026-08-26整合认知负荷与知识迁移的多领域知识追踪
Incorporating Cognitive Load and Knowledge Transfer for Multi-Domain Knowledge Tracing
arXiv 2608.24005 · 2026-08-26AgentSpec：面向大语言模型智能体批量推理的推测解码算法
AgentSpec: Speculative Decoding for Batch Inference of LLM Agents
arXiv 2608.24004 · 2026-08-26每个Plat表示都存在一个正的有界辫子代表元
Every Plat Presentation Admits a Positive Bounded Braid Representative
arXiv 2608.24003 · 2026-08-26陈-西蒙斯修正引力中来自脉冲星周期跃变的引力波宇称破缺应变
Gravitational Wave Parity-Violating Strain from Pulsar Glitches in Chern-Simons Modified Gravity
arXiv 2608.24002 · 2026-08-26具有横向T门的改进量子码
Improved Quantum Codes with Transversal T Gates
arXiv 2608.24000 · 2026-08-26城市的AI权利：公共空间的多元协同设计与治理
The urban right to AI: Pluralistic co-design and governance of public space
arXiv 2608.23999 · 2026-08-26有限总体中多集合分配占用（MAO）的统一精确阶乘矩理论
A Unified Exact Factorial-Moment Theory for Multi-set Allocation Occupancy (MAO) in Finite Populations
arXiv 2608.23998 · 2026-08-26通过Szegő核实现有限剩余卡西米尔与Weil算子的几何化
Geometric Realization of Finite Residue Casimirs and Weil Operators via Szegő Kernels
arXiv 2608.23996 · 2026-08-26用于移动及跨环境场景下鲁棒三维人体姿态估计的物理信息WiFi感知
Physics-Informed WiFi Sensing for Robust 3D Human Pose Estimation in Mobile and Cross-Environment Settings
arXiv 2608.23995 · 2026-08-26通过耦合动力学弥合教师期望与机器人学习之间的差距
Bridging Teacher Expectations and Robot Learning via Coupling Dynamics
arXiv 2608.23994 · 2026-08-26企业MCP网关的混合语义工具发现：架构与实现
Hybrid Semantic Tool Discovery for Enterprise MCP Gateway: Architecture and Implementation
arXiv 2608.23992 · 2026-08-26相对双曲群的相对双精确性及其若干应用
Relative Biexactness for Relative Hyperbolic Groups and Some Applications
arXiv 2608.23991 · 2026-08-26准周期局域化跃迁中动力学临界指数的算术调控
Arithmetic Tuning of Dynamical Critical Exponents in Quasiperiodic Localization Transitions
arXiv 2608.23990 · 2026-08-26有界度图上离散时间与连续时间的碰撞性质
Collision Properties in Discrete and Continuous Time on Bounded-Degree Graphs
arXiv 2608.23989 · 2026-08-26固定T下随机系数动态面板数据模型的估计
Estimation of Random-Coefficient Dynamic Panel Data Models with a Fixed T
arXiv 2608.23988 · 2026-08-26关于非全局Lipschitz系数的随机半线性偏微分方程的最优控制问题
On the Optimal Control Problem of Stochastic Semilinear Partial Differential Equations with Non-Globally Lipschitz Coefficients
arXiv 2608.23985 · 2026-08-26基于单张肖像重建的3D高斯头部的源人脸真实性检测：基准与专用检测器
Source-Face Authenticity Detection for 3D Gaussian Heads Reconstructed from a Single Portrait: A Benchmark and Dedicated Detector
arXiv 2608.23984 · 2026-08-26无传感器的防损伤抓取
Sensorless damage-safe grasping
arXiv 2608.23983 · 2026-08-26加权射影线整体维数界的一个反例
A counterexample to a global-dimension bound for weighted projective lines
arXiv 2608.23981 · 2026-08-26用于Cahn-Hilliard方程的保结构广义可迁移神经网络
Structure-preserving generalized transferable neural networks for the Cahn-Hilliard equation
arXiv 2608.23980 · 2026-08-26规则先于神谕：协商民意调查的可审计、用户可配置论证选择
Rules Before Oracles: Auditable, User-Configurable Argument Selection for Deliberative Polling
arXiv 2608.23979 · 2026-08-26IC-ThermBench：用于通用2.5D/3D-IC热学学习的开放渐进式基准
IC-ThermBench: An Open, Progressive Benchmark for Generalizable 2.5D/3D-IC Thermal Learning
arXiv 2608.23977 · 2026-08-26最大超对称杨-米尔斯理论（$\boldsymbol{\textit{N}=4}$ SYM）中领头对数精度下的三雷吉子交换
Three-Reggeon exchange in $\mathcal{N}=4$ SYM to leading logarithmic accuracy
arXiv 2608.23975 · 2026-08-26用于乳腺超声诊断的通才-专家模型协作的引导与反馈框架
Boot-and-Feedback Framework for Generalist-Expert Model Collaboration in Breast Ultrasound Diagnosis
arXiv 2608.23974 · 2026-08-26乘法上同调Tambara函子
Multiplicatively cohomological Tambara functors
arXiv 2608.23973 · 2026-08-26感知安全的带信号时序逻辑的模型预测路径积分控制
Safety-aware Model Predictive Path Integral Control with Signal Temporal Logic
arXiv 2608.23972 · 2026-08-26正性条件违反下修正处理策略的因果效应：一种部分识别方法
Causal Effects of Modified Treatment Policies under Positivity Violations: A Partial Identification Approach
arXiv 2608.23971 · 2026-08-26Giraffe：一种用于高效平面设计的从隐藏文本表示到视觉嵌入的映射架构
Giraffe: A Mapping Architecture from Hidden Text Representations to Visual Embeddings for Efficient Graphic Design
arXiv 2608.23970 · 2026-08-26探究交互式对话游戏间的知识迁移
Investigating Knowledge Transfer Across Interactive Dialogue Games
arXiv 2608.23969 · 2026-08-26当大语言模型变慢时：环境影响如何调节大学生对大语言模型的使用
When LLMs Slow Down: How Environmental Impacts Mediate University Students' LLM Usage
arXiv 2608.23968 · 2026-08-26Turán $(r+1,r)$-系统最小规模的改进界
An improved bound on the minimum size of Turán $(r+1,r)$-systems
arXiv 2608.23967 · 2026-08-26谁来选择偏好如何聚合？审计基于大语言模型（LLM）的群组推荐中的聚合规则权限
Who Chooses How Preferences Are Aggregated? Auditing Aggregation-Rule Authority in LLM-Based Group Recommendation
arXiv 2608.23966 · 2026-08-26RAGSentinel：用于鲁棒检索增强生成的可验证几何共识
RAGSentinel: Certifiable Geometric Consensus for Robust Retrieval-Augmented Generation
arXiv 2608.23965 · 2026-08-26具有非局部促进与竞争的种群模型中依赖核函数的模式形成
Kernel-Dependent Pattern Formation in a Population Model with Nonlocal Facilitation and Competition
arXiv 2608.23964 · 2026-08-26带有限多个输出函数的Borell–Brascamp–Lieb不等式
Borell--Brascamp--Lieb inequality with finitely many output functions
arXiv 2608.23963 · 2026-08-26更多GPU还是更小的缓存？用于内存受限大语言模型服务的张量并行与KV压缩
More GPUs or a Smaller Cache? Tensor Parallelism versus KV Compression for Memory-Bound LLM Serving
arXiv 2608.23962 · 2026-08-26评估语言模型在跨语言代码功能等价性上的表现
Evaluating Language Models on Cross-Language Code Functional Equivalence
arXiv 2608.23961 · 2026-08-26从依赖不确定性的缺失标签中学习用于半监督分类
Learning from Uncertainty-dependent Missing Labels for Semi-supervised Classification
arXiv 2608.23960 · 2026-08-26NeuronGuard：通过感知消融的安全信号重新分配实现鲁棒的大语言模型安全对齐
NeuronGuard: Robust LLM Safety Alignment via Ablation-Aware Safety Signal Redistribution
arXiv 2608.23959 · 2026-08-26二维邓克尔（Dunkl）牛顿核与格林核的BMO分类
BMO Classification for Two-Dimensional Dunkl Newton and Green Kernels
arXiv 2608.23958 · 2026-08-26基于太极引力波天文台对标量诱导引力波的原初黑洞的投影约束
Projected Constraints on Primordial Black Holes from Scalar-Induced Gravitational Waves with Taiji
arXiv 2608.23957 · 2026-08-26递归智能体推理
Recursive Agentic Reasoning
arXiv 2608.23956 · 2026-08-26自发对称性破缺附近的可调介导相互作用：从汤川势到库仑势及Dzyaloshinskii-Moriya相互作用
Tunable Mediated Interactions Near Spontaneous Symmetry Breaking: From Yukawa to Coulomb and Dzyaloshinskii--Moriya Interactions
arXiv 2608.23954 · 2026-08-26长期分裂的帝国必须统一：三种大语言模型智能体框架的架构趋同
The Empire, Long Divided, Must Unite: Architectural Convergence in Three LLM Agent Harnesses
arXiv 2608.23953 · 2026-08-26
