2026-09论文中文摘要
差异之下：诊断与缓解代码级自主研究循环中的算法模式崩溃
Beneath the Diff: Diagnosing and Mitigating Algorithmic Mode Collapse in Code-Level Autonomous Research Loops
arXiv 2609.00077 · 2026-09-02AI发病率与死亡率：临床AI故障审查框架
AI Morbidity and Mortality: A Framework for Clinical AI Failure Review
arXiv 2609.00076 · 2026-09-02重访预知与自由意志
Foreknowledge and Free Will Revisited
arXiv 2609.00075 · 2026-09-02通过正则化Jackson $q$-积分得到的双基本超几何和
Double basic hypergeometric sums via a regularized Jackson $q$-integral
arXiv 2609.00074 · 2026-09-02MCP客户端在失败后能否决定做什么？仅结果可操作审计
Can MCP Clients Decide What to Do After Failure? A Result-Only Actionability Audit
arXiv 2609.00072 · 2026-09-02AutoXRD：用于粉末衍射分析的自主大语言模型智能体及综合评估
AutoXRD: Autonomous LLM Agents and Comprehensive Evaluation for Powder Diffraction Analysis
arXiv 2609.00070 · 2026-09-02审计自我改进智能体中的控制程序篡改
Auditing Harness Tampering in Self-Improving Agents
arXiv 2609.00069 · 2026-09-02生命算子：一种用于多尺度生命建模的自演化框架
Life Operators: a self-evolving framework for multiscale life modelling
arXiv 2609.00068 · 2026-09-02OCGQuant：面向NVFP4量化的离群值-伴生值分组方法
OCGQuant: Outlier-Companion Grouping for NVFP4 Quantization
arXiv 2609.00066 · 2026-09-02注意力敏感性并非足够：微调下的注意力层级与行为式上下文学习的分离
Attention Sensitivity Is Not Enough: Dissociating Attention-Level and Behavioural In-Context Learning under Fine-Tuning
arXiv 2609.00064 · 2026-09-02基于大型语言模型的医学因果假设验证
Medical Causal Hypothesis Verification with Large Language Models
arXiv 2609.00063 · 2026-09-02RePro：用于可靠评估大语言模型数学问题求解能力的证明验证基准改写框架
RePro: Proof-Verified Benchmark Rewriting for Reliable Evaluation of LLM Mathematical Problem Solving
arXiv 2609.00062 · 2026-09-02智能体支付协议的形式化分析
A Formal Analysis of Agent Payment Protocols
arXiv 2609.00060 · 2026-09-02DISTAL：面向结构无关材料属性预测的蒸馏与自监督预训练框架
DISTAL: Distillation and Self-Supervised Pretraining for Structure-Agnostic Materials Property Prediction
arXiv 2609.00059 · 2026-09-02CUDA-Harness：利用智能体从自然语言生成与优化CUDA内核
CUDA-Harness: Harnessing Agentic CUDA Kernel Generation and Optimization from Natural Language
arXiv 2609.00058 · 2026-09-02ValueGraph：面向情境化用户表示的价值信号引导图预训练
ValueGraph: Value-Signal Guided Graph Pre-training for Contextualized User Representation
arXiv 2609.00057 · 2026-09-02具有狄拉克δ相互作用的等谱势：受约束谱
Isospectral potentials with Dirac delta interaction: Constrained Spectra
arXiv 2609.00056 · 2026-09-02基于大语言模型增强的音频-文本对齐的零样本呼吸声分类
Zero-Shot Respiratory Sound Classification through LLM-Augmented Audio-Text Alignment
arXiv 2609.00055 · 2026-09-02AgentProv：通过工具使用策略探测审计智能体大语言模型API提供商
AgentProv: Auditing Agentic LLM API Providers via Tool-use Policy Probes
arXiv 2609.00052 · 2026-09-02从检测到弃权：通过电路引导的权重缩放实现更安全的大语言模型（LLMs）
From Detection to Refusal: Safer LLMs via Circuit-Guided Weight Scaling
arXiv 2609.00051 · 2026-09-02面向智能体云工程：基于零信任智能体管控器的图与循环工程
Towards Agentic Cloud Engineering: Graph and Loop Engineering with a Zero-Trust Agent Harness
arXiv 2609.00050 · 2026-09-02GUI-CC：将GUI世界模型作为智能体环境对其上下文一致性进行基准测试
GUI-CC: Benchmarking Contextual Consistency of GUI World Models as Agent Environments
arXiv 2609.00048 · 2026-09-02面向多任务图预训练的带有全局上下文的任务特定提示
Task-Specific Prompt with Global Context for Multi-Task Graph Pre-Training
arXiv 2609.00047 · 2026-09-02什么是系统？一般系统论中结构-行为合并的基于交互的解释
What Is a System? An Interaction-Based Account of Structure-Behavior Coalescence in General Systems Theory
arXiv 2609.00043 · 2026-09-02结构-行为融合与传统系统理论的局限
Structure-Behavior Coalescence and the Limits of Traditional Systems Theory
arXiv 2609.00042 · 2026-09-02旋量场精质暗能量的宇宙学扰动及其观测约束
Cosmological Perturbations and Observational Constraints on Spinor Field Quintessence Dark Energy
arXiv 2609.00041 · 2026-09-02麦克斯韦分子玻尔兹曼碰撞算子的福克空间形式化
Fock-Space Formulation of the Boltzmann Collision Operator for Maxwell Molecules
arXiv 2609.00040 · 2026-09-02根基缺陷、维弗里希素数与abc猜想
Radical defects, Wieferich primes, and the $abc$ conjecture
arXiv 2609.00039 · 2026-09-02秩一算子的ρ-数值半径与参数化Buzano型不等式
On the $ρ$-numerical radius of rank-one operators and parametrized Buzano-type inequalities
arXiv 2609.00037 · 2026-09-02SilentProbe：测量作为智能体工具的生产级API中的静默故障
SilentProbe: Measuring Silent Failure in Production APIs Used as Agent Tools
arXiv 2609.00035 · 2026-09-02替代均值迹散度：几何、数据处理与重心
Alternative-mean trace divergences: geometry, data processing, and barycenters
arXiv 2609.00034 · 2026-09-02Lynx2030科学分析组：最终报告
Lynx2030 Science Analysis Group: Final Report
arXiv 2609.00033 · 2026-09-02EULER：通过证据验证的返回机制探索多智能体数学发现中未充分利用的关联
EULER: Exploring Underused Links with Evidence-Checked Return for Multi-Agent Mathematical Discovery
arXiv 2609.00032 · 2026-09-02广义相对论框架下含随时间变化的G和$\Lambda$的宇宙Bianchi I型时空几何：观测方面
Bianchi Type I Space -Time Geometry of the Universe with Time Dependent G and $Λ$ Within the Framework of General Relativity: Observational Aspects
arXiv 2609.00030 · 2026-09-02广义幽灵暗能量对虫洞几何的影响
Influence of Generalized Ghost Dark Energy on Wormhole Geometry
arXiv 2609.00029 · 2026-09-02UI-Venus-2 技术报告
UI-Venus-2 Technical Report
arXiv 2609.00028 · 2026-09-02探究F(R)引力理论中新黑洞解的热力学与光子性质
Exploring thermodynamic and photonic properties of new black hole solutions in F(R) gravity theory
arXiv 2609.00027 · 2026-09-02在边缘设备上针对不同感知模态的脉冲神经网络基准测试
Benchmarking spiking neural networks across sensing modalities on edge devices
arXiv 2609.00026 · 2026-09-02具有修正温度的四维爱因斯坦-高斯-博内特黑洞的热力学
Thermodynamics of four-dimensional Einstein-Gauss-Bonnet black holes with a modified temperature
arXiv 2609.00025 · 2026-09-02有限五智能体ABCW模型中单步预测的精确最小场划分
Exact Minimum Field Partition for One-Step Prediction in a Finite Five-Agent ABCW Model
arXiv 2609.00024 · 2026-09-02ES-AHD：一种用于自动启发式设计的进化策略框架
ES-AHD: An Evolution Strategy Framework for Automatic Heuristic Design
arXiv 2609.00023 · 2026-09-02基于DESI-DR2观测检验f(R,T)引力框架下的体黏度
Testing the bulk viscosity within the f(R,T) gravity in light of DESI-DR2 observations
arXiv 2609.00022 · 2026-09-02Calabi-Yau三维奇点及其通用被单
Calabi-Yau Threefold Singularities and their Universal Quilts
arXiv 2609.00021 · 2026-09-02离散电磁学中的陈-西蒙斯涨落与信息几何
Chern--Simons Fluctuations and Information Geometry in Discrete Electromagnetism
arXiv 2609.00020 · 2026-09-02SCAFFOLD：包含图表问答与思维链推理轨迹的大规模计算机科学研究图表结构化数据集
SCAFFOLD: A Large-Scale Structured Dataset of Computer Science Research Figures with Diagram QA and Chain-of-Thought Reasoning Traces
arXiv 2609.00018 · 2026-09-02f(R,L_m)引力中太初核合成的可行性
Viability of Big Bang Nucleosynthesis in $f(R,L_m)$ Gravity
arXiv 2609.00017 · 2026-09-02带刺猬标量毛的几何正则黑洞的奇宇称扰动
Odd-Parity Perturbations of Geometrically Regular Black Holes with Hedgehog Scalar Hair
arXiv 2609.00016 · 2026-09-02面向个性化对齐与多视角推理的、来自真实场景的行为基础用户画像
Behaviorally Grounded User Profiles from the Wild for Personalized Alignment and Multi-Perspective Reasoning
arXiv 2609.00014 · 2026-09-02在双通道噪声数据中高效搜索小信号：大数据观测中的一项挑战
Efficient searches of small signals across two-channel noisy data: a challenge in Big Data observations
arXiv 2609.00013 · 2026-09-02大语言模型中的长程状态跟踪：通过深度依赖工具调用序列执行MD5
Long-Horizon State Tracking in LLMs: Executing MD5 through a Deep Sequence of Dependent Tool Calls
arXiv 2609.00012 · 2026-09-02
