2026-08论文中文摘要

• RefVideo-6M：一个用于教学视频编辑的可靠基于参考的数据集

  RefVideo-6M: A Reliable Reference-Based Dataset for Instructional Video Editing

  arXiv 2608.26101 · 2026-08-27

• 基于朗之万动力学的集成硬件退火算法用于伊辛机

  Integrated Hardware Annealing based on Langevin Dynamics for Ising Machines

  arXiv 2608.26100 · 2026-08-27

• 探测本地泡附近中性云的性质：将三维尘埃与中性氢吸收联系起来

  Probing the Properties of Neutral Clouds in the Vicinity of the Local Bubble: Linking three-dimensional Dust with Neutral Hydrogen Absorption

  arXiv 2608.26098 · 2026-08-27

• 扭秤作为半经典引力的运作探针：匹配滤波界限、扭矩扩散约束与量子噪声基准

  Torsion balances as operational probes of semiclassical gravity: Matched-filter bounds, torque-diffusion constraints, and quantum-noise benchmarks

  arXiv 2608.26096 · 2026-08-27

• 面向多模态无监督持续后训练的视觉依赖感知框架

  A Visual Dependence-Aware Framework for Multimodal Unsupervised Continual Post-Training

  arXiv 2608.26095 · 2026-08-27

• MyoMechanix：基于生物力学的可组合技能活动理解与指导

  MyoMechanix: Biomechanically-Grounded Compositional Skilled Activity Understanding and Coaching

  arXiv 2608.26094 · 2026-08-27

• 面向小区边缘功率控制的智能体自动研究：从根本上重新定义研究者的角色

  Agentic Autoresearch for Cell-Edge Power Control: Radically Redefining the Researcher's Role

  arXiv 2608.26093 · 2026-08-27

• 偶维球面上各向同性正定函数的最优可微性

  Optimal differentiability of isotropic positive definite functions on even-dimensional spheres

  arXiv 2608.26092 · 2026-08-27

• PlanSightRAG：面向民用标准图纸的自动化问答与合规检查的视觉优先多模态检索增强生成模型

  PlanSightRAG: A Visual-First Multimodal RAG for Automating Question Answering and Compliance Checking for Civil Standard Plans

  arXiv 2608.26091 · 2026-08-27

• 利用稀疏自编码器在中微子基础模型中寻找和使用可解释隐变量

  Finding and using interpretable latents in a neutrino foundation model with sparse autoencoders

  arXiv 2608.26090 · 2026-08-27

• 从生产到验证：人工智能如何削弱自由职业者的技能

  From Producing to Validating: How AI Is Deskilling Freelancers

  arXiv 2608.26089 · 2026-08-27

• 二次引力中的大质量鬼场禁闭、偶极方程与多极态

  Massive Ghost Confinement, Dipole Equation and Multipole States in Quadratic Gravity

  arXiv 2608.26085 · 2026-08-27

• 通过锥形框架实现局部ic Esakia对偶性

  Localic Esakia Duality via Conic Frames

  arXiv 2608.26084 · 2026-08-27

• 无后选择的非厄米物理量子模拟：作为耗散动力学相变的Yang-Lee临界性

  Yang-Lee Criticality as a Dissipative Dynamical Phase Transition: Quantum Simulation of non-Hermitian Physics without Post-selection

  arXiv 2608.26082 · 2026-08-27

• SwarmWorld：语言模型智能体社会中的标记介导技术演化

  SwarmWorld: Stigmergic technological evolution in societies of language-model agents

  arXiv 2608.26081 · 2026-08-27

• 曲面上线丛的负有效除子与Bridgeland稳定性

  Negative Effective Divisors and Bridgeland Stability of Line Bundles on Surfaces

  arXiv 2608.26080 · 2026-08-27

• 广义锥猜想，第一部分：超越卡拉比-丘流形

  Generalised Cone Conjecture, I: Beyond Calabi--Yau

  arXiv 2608.26079 · 2026-08-27

• 宇称反常作为含无质量狄拉克费米子的模对易子

  Parity Anomaly as Modular Commutator with Massless Dirac Fermion

  arXiv 2608.26078 · 2026-08-27

• 广延空气簇射的剖析：构建最优的辐射计算方法

  Anatomy of an extensive air shower: building an optimal radio emission calculation

  arXiv 2608.26077 · 2026-08-27

• 基于李群约束MeanFlow的快速生成式抓取

  Fast Generative Grasping via Lie Group-Constrained MeanFlow

  arXiv 2608.26076 · 2026-08-27

• 认知网络、集体误解与社会知识的操纵

  Epistemic Networks, Collective Misperception, and the Manipulation of Social Knowledge

  arXiv 2608.26075 · 2026-08-27

• 承诺前的门控：预测意图分歧以防止自动驾驶中交互后的决策失败

  Gating Before Commitment: Anticipating Intent Divergence to Prevent Post-Interaction Decision Failures in Autonomous Driving

  arXiv 2608.26074 · 2026-08-27

• 等离子体二聚体的光消光与近场特性：粒子形状和间距的作用

  Optical extinction and near-field properties of plasmonic dimers: Role of particle shape and separation

  arXiv 2608.26073 · 2026-08-27

• 从车队到实验室：重新审视工业级Rowhammer缓解方案的安全性与复杂性

  From Fleet to Lab: Revisiting the Security and Complexity of Industrial Rowhammer Mitigation

  arXiv 2608.26072 · 2026-08-27

• 一维近藤金属CeCo₂Ga₈中的非公度自旋涨落

  Incommensurate spin fluctuations in one-dimensional Kondo metal CeCo2Ga8

  arXiv 2608.26071 · 2026-08-27

• 用于高效测试时缩放的前缀滑动

  Prefix Sliding for efficient test-time scaling

  arXiv 2608.26070 · 2026-08-27

• 所有一次扩展的三维拓扑量子场论（TQFT）均为Reshetikhin–Turaev理论

  All once-extended 3D TQFTs are Reshetikhin--Turaev theories

  arXiv 2608.26068 · 2026-08-27

• StreamPI：面向视觉-语言-动作模型的流式多模态时序建模

  StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models

  arXiv 2608.26067 · 2026-08-27

• VirTooS：用于自主移动机器人车队管理的ROS 2-Unity虚拟化工具包

  VirTooS: A ROS 2 - Unity Virtualization Toolkit for Fleet Management of Autonomous Mobile Robots

  arXiv 2608.26066 · 2026-08-27

• 实验室地震证实了摩擦滑动脉冲理论

  Lab earthquakes confirm the theory of frictional slip pulses

  arXiv 2608.26065 · 2026-08-27

• AI驱动CFD的神经网络与降阶建模工作流：快速响应面、降阶动力学及横流喷流示例

  Neural-Network and Reduced-order Modeling Workflows for AI-Driven CFD: Fast Response Surfaces, Reduced Dynamics and Jet in Cross-flow Examples

  arXiv 2608.26064 · 2026-08-27

• 狄拉克型左右对称理论中的强CP与PMNS相位

  Strong CP and the PMNS Phase in Dirac Left-Right Symmetry

  arXiv 2608.26063 · 2026-08-27

• 任意维度下噪声幺正信道的渐近最优纯化

  Asymptotically optimal purification of noisy unitary channels in any dimension

  arXiv 2608.26061 · 2026-08-27

• 微调Whisper以实现巴尼瓦语的自动语音识别：一项初步研究

  Fine-Tuning Whisper for Automatic Speech Recognition in Baniwa: A Preliminary Study

  arXiv 2608.26060 · 2026-08-27

• 阿达马空间中等于及高于渐近秩的线性等周填充不等式

  Linear isoperimetric filling inequalities in Hadamard spaces at and above the asymptotic rank

  arXiv 2608.26059 · 2026-08-27

• 单一策略，多种具现体：面向异构具身操纵的以相机为中心的统一动作几何预训练

  One Policy, Many Embodiments: Unified Camera-Centric Action Geometry Pre-training for Heterogeneous Embodied Manipulation

  arXiv 2608.26058 · 2026-08-27

• 氮化镓中室温量子发射器的深度控制

  Depth Control of Room-Temperature Quantum Emitters in Gallium Nitride

  arXiv 2608.26057 · 2026-08-27

• 为机械工程师提供智能工具：一个基于项目的热工领域人工智能教育课程

  Giving Mechanical Engineers Intelligent Tools: A Project-Based AI Education Curriculum in Thermal Engineering

  arXiv 2608.26056 · 2026-08-27

• 用于暗孤子的数值直接散射变换

  Numerical Direct Scattering Transform for Dark Solitons

  arXiv 2608.26054 · 2026-08-27

• R^3：通过强化学习训练机器人以自然语言进行推理

  $R^3$: Training Robots to Reason in Natural Language via Reinforcement Learning

  arXiv 2608.26053 · 2026-08-27

• LoRA需要多少秩？Transformer注意力的秩-误差边界

  How Much Rank Does LoRA Need? Rank-Error Bounds for Transformer Attention

  arXiv 2608.26052 · 2026-08-27

• 人类专业知识的价值

  The Value of Human Expertise

  arXiv 2608.26051 · 2026-08-27

• RTLGuard：一种针对中毒RTL代码生成模型的轻量级师生防御方法

  RTLGuard: A Lightweight Teacher-Student Defense for Poisoned RTL Code Generation Models

  arXiv 2608.26049 · 2026-08-27

• 用于电推进应用的增材制造金属合金的实验表征

  Experimental Characterization of Additively Manufactured Metallic Alloys for Electric Propulsion Applications

  arXiv 2608.26048 · 2026-08-27

• 用于Cahn-Hilliard-Navier-Stokes方程的一族二阶线性无条件稳定方法

  A family of second order, linear, unconditionally stable methods for the Cahn-Hilliard-Navier-Stokes equations

  arXiv 2608.26046 · 2026-08-27

• 多端YBa$_2$Cu$_3$O$_{7-δ}$结中氧化学计量比的电学调控

  Electrical manipulation of oxygen stoichiometry in multiterminal YBa$_2$Cu$_3$O$_{7-δ}$ junctions

  arXiv 2608.26045 · 2026-08-27

• 多项式扩张不保持强有限型性质

  Polynomial extensions do not preserve the strong finite type property

  arXiv 2608.26044 · 2026-08-27

• 鲁棒CurveMoE：通过模式连通性实现混合专家模型的多范数对抗防御

  Robust CurveMoE: Multi-Norm Adversarial Defense for Mixture-of-Experts Models via Mode Connectivity

  arXiv 2608.26043 · 2026-08-27

• 完备渐近锥形流形上的$G_2$-拉普拉斯流

  $G_2$-Laplacian flow on complete asymptotically conical manifolds

  arXiv 2608.26042 · 2026-08-27

• 狄利克雷级数哈代空间上有限素数复合算子的临界范数轮廓

  Critical Norm Profiles for Finite-Prime Composition Operators on the Hardy Space of Dirichlet Series

  arXiv 2608.26041 · 2026-08-27