2026-08论文中文摘要
用于推理的思维级束搜索
Thought-Level Beam Search for Reasoning
arXiv 2608.08020 · 2026-08-12努力而非明智：推理模型无法在不同问题间合理分配测试时计算资源
Thinking Hard, Not Smart: Reasoning Models Fail to Ration Test-Time Compute Across Questions
arXiv 2608.07968 · 2026-08-12量子中国剩余时钟
Quantum Chinese Remainder Clock
arXiv 2608.07938 · 2026-08-12面向策略内自蒸馏的自适应监督锚定
Adaptive Supervised Anchoring for On-Policy Self-Distillation
arXiv 2608.07935 · 2026-08-12TongGuOCR：面向中文历史文献的布局感知与 token 增强 OCR 框架
TongGuOCR: A Layout-Aware and Token-Augmented OCR MLLM for Chinese Historical Documents
arXiv 2608.07917 · 2026-08-12DeCo：通过解耦与重耦实现零样本工业异常生成
DeCo: Zero-Shot Industrial Anomaly Generation through Decoupling and Recoupling
arXiv 2608.07904 · 2026-08-12UniScale：任意尺度的工业异常生成
UniScale: Arbitrary-Scale Industrial Anomaly Generation
arXiv 2608.07864 · 2026-08-12关于巴内基定理的一个注记
A remark on Banecki's theorem
arXiv 2608.07843 · 2026-08-12用于CubeSat高光谱成像有效载荷的紧凑型单片折反式透镜设计
Design of a Compact Monolithic Catadioptric Lens for CubeSat Hyperspectral Imaging Payloads
arXiv 2608.07784 · 2026-08-12面向卫星巡检任务的信息感知模型预测控制
Information-Aware Model Predictive Control for Satellite Inspection
arXiv 2608.07765 · 2026-08-12通过CIV吸收线探测当前时代亮星系的星系际介质
Circumgalactic CIV Absorption Found Only Within 30 Degrees of the Minor Axis of Present-Epoch Galaxies
arXiv 2608.07752 · 2026-08-12HSMLA：用于高效视觉Transformer的分层Softmax多尺度线性注意力
HSMLA: Hierarchical Softmax Multi-scale Linear Attention for Efficient Vision Transformers
arXiv 2608.07616 · 2026-08-12MasDrift：跨多智能体架构的授权保留基准测试
MasDrift: Benchmarking Authorization Preservation Across Multi-Agent Architectures
arXiv 2608.07556 · 2026-08-12立场：我们需要针对人类福祉优化的大语言模型
Position: We Need Large Language Models Optimized For Our Well-Being
arXiv 2608.07505 · 2026-08-12用于支持智能家居中独立生活的人体活动识别的IMU数据集（IMU-HAR-IL）
An IMU Dataset for Human Activity Recognition to Support Independent Living in Smart Homes (IMU-HAR-IL)
arXiv 2608.07502 · 2026-08-12EliSeg：面向报告驱动的异常分割的经验证目标构建
EliSeg: Verified Target Construction for Report-Grounded Abnormality Segmentation
arXiv 2608.07299 · 2026-08-12“操作员，您能听到我吗？”通往UNISOC基带的可靠线路
"Operator, can you hear me?" A Faithful Line into the UNISOC Baseband
arXiv 2608.07143 · 2026-08-12I型 seesaw 模型中通过共振序贯主导和 TBC3 混合实现轻子生成
Leptogenesis via Resonant Sequential Dominance and TBC3 Mixing in a Type-I Seesaw Model
arXiv 2608.06983 · 2026-08-12TRIBE：通过通信行为集成预测团队绩效
TRIBE: Predicting Team Performance via Communication Behavior Ensembles
arXiv 2608.06926 · 2026-08-12弱监督学习的最新进展：新的监督范式、假设松弛与实用解决方案
Recent advances in weakly supervised learning: New supervision paradigms, assumption relaxations, and practical solutions
arXiv 2608.06896 · 2026-08-12DAEP：面向医学视频语料时序答案 grounding 的难度感知证据规划
DAEP: Difficulty-Aware Evidence Planning for Medical Video Corpus Temporal Answer Grounding
arXiv 2608.06869 · 2026-08-12亚真空干扰用于安全通信
Sub-Vacuum Jamming for Secure Communication
arXiv 2608.06720 · 2026-08-12策略掩码私有专家：稀疏混合专家（MoE）模型中的可审计可逆能力访问控制
Policy-Masked Private Experts: Auditable and Reversible Capability Access Control in Sparse MoE Models
arXiv 2608.06690 · 2026-08-12铺砌、填充与施瓦茨类加伯窗的存在性
Tilings, packings, and the existence of Schwartz-class Gabor windows
arXiv 2608.06679 · 2026-08-12带有少量信息的拍卖设计
Auction Design with a Bit of Information
arXiv 2608.06623 · 2026-08-12基于采样的量子对角化方法估算HeH$^{+}$、ArH$^{+}$和H$_2$O的基态能量
Ground-State Energy Estimation of HeH$^{+}$, ArH$^{+}$, and H$_2$O via Sample-Based Quantum Diagonalization
arXiv 2608.06415 · 2026-08-12具有Hölder扩散系数的随机时间非局部电报方程的适定性：遗传相空间提升与新型广义耦合方法
Well-posedness of stochastic time-nonlocal telegraph equations with Hölder diffusion coefficient: hereditary phase-space lifting and novel generalized coupling method
arXiv 2608.06237 · 2026-08-12无限群上的逆模糊映射
Inverse ambiguous maps on infinite groups
arXiv 2608.06191 · 2026-08-12面向搜索智能体的上下文信息策略优化
Contextual Information Policy Optimization for Search Agents
arXiv 2608.06128 · 2026-08-12ECHO：具备时序记忆、安全护栏与语音评估功能的本地可部署智能体式健康助手
ECHO: A Locally-Deployable Agentic Health Assistant with Temporal Memory, Safety Guardrails, and Speech Assessment
arXiv 2608.06110 · 2026-08-12月球极地冰质月壤与可能的月流
The Lunar Polar Icy Regolith and Possible Moonflows
arXiv 2608.06087 · 2026-08-12提交前的证据锁定：冻结界面会降低“大模型作为评判者”的评估效果
Evidence Lock Before Commitment: A Frozen Interface Degrades LLM-as-Judge Evaluation
arXiv 2608.05353 · 2026-08-12人类认知与行为的小型基础模型
Small Foundation Models of Human Cognition and Behaviour
arXiv 2608.05224 · 2026-08-12常加权平均曲率曲面的亚历山德罗夫定理
Alexandrov Theorem for constant weighted mean curvature surfaces
arXiv 2608.05119 · 2026-08-12CIFAR-10上的规范联合能量基模型：预测-校正采样器与SGLD采样器的失效模式及实际不可区分性
Comparing SGLD and a fixed-noise Predictor-Corrector adaptation in canonical Joint Energy-Based Models on CIFAR-10
arXiv 2608.05025 · 2026-08-12迈向MEV市场中的去中心化搜索者竞争
Towards Decentralized Searcher Competition in MEV Markets
arXiv 2608.05011 · 2026-08-12量子比特的互补量子关联具有普适性
Complementary Quantum Correlations Are Universal for Qubits
arXiv 2608.04916 · 2026-08-12双有理动机同伦范畴的概型函子性
Schematic Functorialities of Birational Motivic Homotopy Categories
arXiv 2608.04793 · 2026-08-12冷里德伯原子的杂散电场光电离补偿
Photo-ionization Compensation of Stray Electric Fields for Cold Rydberg Atoms
arXiv 2608.04660 · 2026-08-12从低质量双轻子中分离出强子再散射的物态方程动力学
Separating Equation-of-State Dynamics from Hadronic Rescattering in Low-Mass Dileptons
arXiv 2608.04598 · 2026-08-12三部染色临界图的边谱过饱和
Edge-spectral supersaturation for tripartite color-critical graphs
arXiv 2608.04485 · 2026-08-12耦合漂移Monge-Ampère方程的刘维尔型定理
Liouville-type theorems for coupled-drift Monge-Ampère equations
arXiv 2608.04478 · 2026-08-12半导体太阳能电池材料与器件中的量子光电子学
Quantum optoelectronics in semiconductor solar cell materials and devices
arXiv 2608.04328 · 2026-08-12实验Kuramoto平台
Experimental Kuramoto Platform
arXiv 2608.04119 · 2026-08-12地理空间基础模型的可用性如何？对89个模型的系统评估
How Usable Are Geospatial Foundation Models? A Systematic Evaluation of 89 Models
arXiv 2608.03804 · 2026-08-12集成宇宙学记忆：一种探测暗能量的暗信号源方法
Integrated cosmological memory: A dark-siren method to probe dark energy
arXiv 2608.03739 · 2026-08-12量子杂质作为二维玻色气体有限温度涨落的探针
Quantum Impurities as Probes of Finite-Temperature Fluctuations in Two-Dimensional Bose Gases
arXiv 2608.03665 · 2026-08-12CastFSR：用于上下文感知时间序列预测的快慢反思智能体推理框架
CastFSR: A Fast--Slow--Reflect Agentic Reasoning Framework for Context-Aware Time Series Forecasting
arXiv 2608.03031 · 2026-08-12极值正则黑洞的形成
Formation of extremal regular black holes
arXiv 2608.02908 · 2026-08-12科学实验场景建模：数据集与模型
Modeling Scientific Experiment Scenes: Dataset and Model
arXiv 2608.02892 · 2026-08-12
