2026-08论文中文摘要
面向低轨（LEO）卫星网络中基于强化学习（RL）的路由的大语言模型（LLM）驱动的自动奖励设计
LLM-Driven Automated Reward Design for Reinforcement Learning-Based Routing in LEO Satellite Networks
arXiv 2608.01649 · 2026-08-04大规模高性能计算系统硬件错误预测技术评估
Evaluating Forecasting Techniques for Hardware Errors on a Large-scale HPC System
arXiv 2608.01648 · 2026-08-04NEXUS：红矮点与蓝活跃星系核在2≲z≲6时的光谱变异性
NEXUS: Spectral Variability of Little Red Dots and Blue Active Galactic Nuclei at $2 \lesssim z \lesssim 6$
arXiv 2608.01647 · 2026-08-04Spike-HTR：用于手写文本识别的脉冲神经Transformer
Spike-HTR: Spiking Neural Transformer for Handwritten Text Recognition
arXiv 2608.01646 · 2026-08-04GISAgentBench：面向GIS任务的、由从业者提供的用于评估大语言模型智能体的基准测试
GISAgentBench: A Practitioner-Sourced Benchmark for Evaluating LLM Agents on GIS Tasks
arXiv 2608.01645 · 2026-08-04CRAFT：面向视觉语言模型的基于视频令牌递归自适应融合的压缩方法
CRAFT: Compression via Recursive Adaptive Fusion of Video Tokens for Vision-Language Models
arXiv 2608.01644 · 2026-08-04StreamTalk：基于关键姿态锚定的流式伴随语音手势生成
StreamTalk: Streaming Co-Speech Gesture Generation with Key-Pose Anchoring
arXiv 2608.01643 · 2026-08-04用于恶意软件预测的机器学习算法性能分析
Performance analysis of Machine learning algorithms for predicting malware
arXiv 2608.01642 · 2026-08-04AI辅助的需求启发访谈脚本管理
AI-assisted Script Management for Requirements Elicitation Interviews
arXiv 2608.01640 · 2026-08-04变异规避：基于知识驱动多智能体编排的自主端点规避
Mutate to Bypass: Autonomous Endpoint Evasion via Knowledge-Driven Multi-Agent Orchestration
arXiv 2608.01639 · 2026-08-04面向高效自我中心 grounding 的动态分辨率路由
Dynamic Resolution Routing for Efficient Egocentric Grounding
arXiv 2608.01638 · 2026-08-04Salami攻击：针对OpenClaw的隐蔽合谋式内存投毒
Salami Attack: Stealthy Collusive Memory Poisoning against OpenClaw
arXiv 2608.01637 · 2026-08-04用于增强多智能体轨迹规划的前向-逆向动态博弈框架
A Forward-Inverse Dynamic Game Framework for Enhanced Multi-Agent Trajectory Planning
arXiv 2608.01636 · 2026-08-04通过空间-光谱视觉锚学习缓解多模态大语言模型（MLLMs）中的视觉退化
Mitigating Visual Degradation in MLLMs via Spatial-Spectral Visual Anchor Learning
arXiv 2608.01635 · 2026-08-04利用引力波探测器网络确定核心坍缩超新星的参数估计范围
Parameter Estimation Horizon of Core-Collapse Supernovae with a Network of Gravitational-Wave Detectors
arXiv 2608.01634 · 2026-08-04GraphIR：面向大语言模型引导的神经架构演化的架构级搜索状态
GraphIR: Architecture-Level Search States for LLM-Guided Neural Architecture Evolution
arXiv 2608.01633 · 2026-08-04准确率等同于证据吗？KV缓存压缩下的推理忠实性
Does Accuracy Equal Evidence? Reasoning Faithfulness under KV Cache Compression
arXiv 2608.01631 · 2026-08-04RING：用于持续大规模知识注入的检索内化生成
RING: Retrieval-Internalized Generation for Continual Large-Scale Knowledge Injection
arXiv 2608.01630 · 2026-08-04在非母语日语的语言态度上实现人类与大语言模型（LLM）的对齐
Human-LLM Alignment in Language Attitudes Toward Non-Native Japanese
arXiv 2608.01629 · 2026-08-04具有生物记忆项的粒子与动力学系统中的渐近群聚行为
Asymptotic Flocking Behavior in Particle and Kinetic Systems with biological Memory Term
arXiv 2608.01627 · 2026-08-04QWRF-Net：用于短期降水临近预报的带整流流的量子-小波框架
QWRF-Net: A Quantum-Wavelet Framework with Rectified Flow for Short-Term Precipitation Nowcasting
arXiv 2608.01626 · 2026-08-04并非维度，而是范数：无梯度语言模型权重扰动的关键因素
Not the Dimension, the Norm: What Matters in Gradient-Free Weight Perturbation of Language Models
arXiv 2608.01624 · 2026-08-04SAGE：面向试次内切换的脑电引导软门控目标说话人提取方法
SAGE: Switch-Aware EEG-Guided Soft Gating for Target Speaker Extraction with In-Trial Switching
arXiv 2608.01623 · 2026-08-04SMM Transformer：利用脉冲神经网络实现多模态任务
SMM Transformer: Leveraging Spiking Neural Networks for Multimodal Tasks
arXiv 2608.01622 · 2026-08-04日冕物质抛射（CME）及其驱动的激波因与环境太阳风相互作用产生的形变：I. 基于锥模型的模拟
Deformation of the CME and CME-driven shock due to interaction with the ambient solar wind: I. Modelling with Cone Model
arXiv 2608.01621 · 2026-08-04通过心电引导跨模态预训练从智能手表光体积描记术衍生的心龄作为血管衰老的数字生物标志物
Smartwatch Photoplethysmography-Derived Heart Age via ECG-Guided Cross-Modal Pretraining as a Digital Biomarker of Vascular Aging
arXiv 2608.01620 · 2026-08-04当记忆更新但行为未更新时：修复个性化智能体响应中的隐式陈旧依赖
When Memory Updates but Behavior Does Not: Repairing Implicit Stale Dependencies in Personalized Agent Responses
arXiv 2608.01619 · 2026-08-04SGAD：面向鲁棒性基于EEG的听觉注意力切换解码的状态引导自适应决策框架
SGAD: A State-Guided Adaptive Decision Framework for Robust EEG-Based Auditory Attention Switch Decoding
arXiv 2608.01618 · 2026-08-04基于少量子比特测量的近最优量子计量学
Near-optimal quantum metrology with few-qubit measurements
arXiv 2608.01617 · 2026-08-04基于极小极大与后验匹配的在线算法
Online Algorithms via Minimax and Posterior Matching
arXiv 2608.01616 · 2026-08-04线性多时间尺度保留作为内存高效的视觉-语言桥梁
Linear Multi-Timescale Retention as a Memory-Efficient Vision-Language Bridge
arXiv 2608.01614 · 2026-08-04从“病毒式”到“虚空”：用于识别“ rug pull”的多维行为与合约分析
From Viral to Void: Multi-Dimensional Behavioral and Contractual Analysis for Rug Pull Identification
arXiv 2608.01609 · 2026-08-04半正矩阵的可逆线性保持映射——一种无维度方法
Invertible linear preservers of semipositive matrices - a dimension free approach
arXiv 2608.01608 · 2026-08-04AI理财建议：供给、需求与生命周期影响
AI Financial Advice: Supply, Demand, and Life Cycle Implications
arXiv 2608.01607 · 2026-08-04ETHead：从语音生成富有表现力的3D面部动画与头部运动
ETHead: Generating Expressive 3D Facial Animation and Head Movement from Speech
arXiv 2608.01605 · 2026-08-04在办公工作上进行后训练可提升软件工程能力：跨域迁移的行为视角
Post-Training on Office Work Improves Software Engineering: A Behavioral Account of Cross-Domain Transfer
arXiv 2608.01604 · 2026-08-04用于4D-STEM中稀疏应变场重建的物理信息神经网络
Physics-Informed Neural Networks for Sparse Strain-Field Reconstruction in 4D-STEM
arXiv 2608.01601 · 2026-08-04面向建筑施工任务执行的人形机器人感知与动作系统
Perception-and-action system for humanoid robot task execution in construction
arXiv 2608.01600 · 2026-08-04用于波动率预测的潜在制度偏差审计
Latent-Regime Bias Auditing for Volatility Forecasting
arXiv 2608.01599 · 2026-08-04PICTURE：通过揭示而非隐藏角色的知识缺失来增强大语言模型的心智理论能力
PICTURE: Enhancing Theory-of-Mind in Large Language Models by Revealing, Not Hiding, Characters' Lack of Knowledge
arXiv 2608.01598 · 2026-08-04HindSearch：面向搜索增强型强化学习的轨迹级事后批判方法
HindSearch: Trajectory-Level Hindsight Critique for Search-Augmented Reinforcement Learning
arXiv 2608.01597 · 2026-08-04超越K-理论：双曲能带理论中的几何与全纯性
Beyond $K$-Theory: Geometry and Holomorphy in Hyperbolic Band Theory
arXiv 2608.01596 · 2026-08-04代数群的有限阿贝尔子群
Finite abelian subgroups of algebraic groups
arXiv 2608.01595 · 2026-08-04潜思维信用：面向潜推理的多答案信用分配
Latent Thought Credit: Multi-Answer Credit Assignment for Latent Reasoning
arXiv 2608.01593 · 2026-08-04大型语言模型在软件工程领域生成分类体系的表现如何？一个多视角评估框架
How Well Do LLMs Generate Taxonomies in the SE Domain? A Multi-perspective Evaluation Framework
arXiv 2608.01592 · 2026-08-04用于无需人工干预的广告市场的一次性定价
One-Shot Pricing for Hands-Off-the-Wheel Advertising Markets
arXiv 2608.01591 · 2026-08-04基于参数模型的高维协方差矩估计方法
Method of Moments Estimation of High-Dimensional Covariance Using a Parametric Model
arXiv 2608.01590 · 2026-08-04更多特权信息更好吗？自蒸馏推理中的解轨迹到问题解决结构
Is More Privileged Information Better? From Solution Traces to Problem-Solving Structure in Self-Distilled Reasoning
arXiv 2608.01589 · 2026-08-04D^2-4DGS：双深度引导的稀疏相机四维高斯溅射
D^2-4DGS: Dual-Depth Guided Sparse-Camera 4D Gaussian Splatting
arXiv 2608.01588 · 2026-08-04标签定义时间尺度：时间聚合学习的特质-状态极限
The Label Defines the Timescale: Trait-State Limits of Temporal-Aggregate Learning
arXiv 2608.01587 · 2026-08-04
