2026-08论文中文摘要
有限宽度的神经网络场论
Neural Network Field Theory at Finite Width
arXiv 2608.21588 · 2026-08-25一种无需精细调节即可产生长时标的极简机制
A minimal mechanism to generate long timescales without fine tuning
arXiv 2608.21586 · 2026-08-25Asayama-Matsumoto猜想与改进的偏差界
The Asayama-Matsumoto conjecture and a refined discrepancy bound
arXiv 2608.21585 · 2026-08-25基于大语言模型的数据驱动动态算法调度
Data-Driven Dynamic Algorithm Dispatch with Large Language Models
arXiv 2608.21584 · 2026-08-25用于口腔癌筛查的鲁棒轻量级深度学习模型
Robust Lightweight Deep Learning Models for Oral Cancer Screening
arXiv 2608.21583 · 2026-08-25读取房间：循环世界模型状态中的隐式困惑编码
Reading the Room: Implicit Confusion Encoding in Recurrent World Model States
arXiv 2608.21582 · 2026-08-25基于反例的排序
Sorting from Counterexamples
arXiv 2608.21579 · 2026-08-25非可积系统中测量诱导相变的标度与纠缠
Scaling vs entanglement in measurement-induced phase transition for non-integrable systems
arXiv 2608.21578 · 2026-08-25锚定偏差：一种针对持续学习下多模态大语言模型（MLLMs）的持久公平性后门攻击
Anchoring Bias: A Persistent Fairness Backdoor Attack against MLLMs under Continual Learning
arXiv 2608.21577 · 2026-08-25亚秒差距尺度下M87喷流中的相对论性团块特征
Relativistic blob signatures in the M87 jet at sub-parsec scales
arXiv 2608.21576 · 2026-08-25液压网络不可压缩流动的数据驱动计算框架
A Data-Driven Computational Framework for Incompressible Flow in Hydraulic Networks
arXiv 2608.21575 · 2026-08-25鲁棒（整数）线性规划的复杂性图景
The complexity landscape of robust (integer) linear programming
arXiv 2608.21574 · 2026-08-25非阿贝尔大筛法与阿廷原根猜想
A non-abelian large sieve and Artin's primitive root conjecture
arXiv 2608.21573 · 2026-08-25为仿真到真实性能证书下注
Betting for Sim-to-Real Performance Certificates
arXiv 2608.21572 · 2026-08-25拓展早期诊断的视野：利用视觉Transformer进行肺癌预测
Extending the Horizon of Early Diagnosis: Lung Cancer Prediction with Vision Transformers
arXiv 2608.21571 · 2026-08-25适用于CPU可部署安全分类的可复现、许可感知蒸馏方案
A Reproducible, License-Aware Distillation Recipe for CPUDeployable Safety Classification
arXiv 2608.21570 · 2026-08-25用参数驱动腔模拟非马尔可夫系统-库动力学
Emulating non-Markovian system-bath dynamics with parametrically driven cavities
arXiv 2608.21569 · 2026-08-25网络多智能体线性二次差分博弈中带输出反馈的保证成本均衡
Output Feedback Guaranteed Cost Equilibrium in Networked Multi-Agent Linear-Quadratic Difference Games
arXiv 2608.21568 · 2026-08-25$\boldsymbol{\frak{gl}_N}$网络范畴中的6价顶点及其范畴化
6-valent vertex in the $\mathfrak{gl}_N$ web category and its categorification
arXiv 2608.21566 · 2026-08-25基于Logistic-Lerch近似的Nakagami-m衰落信道上的平均有限块长分组错误率
Average Finite-Blocklength Packet Error Rate over Nakagami-$m$ Fading via a Logistic--Lerch Approximation
arXiv 2608.21564 · 2026-08-25基于可解释人工智能的1型糖尿病夜间低血糖预防决策支持
Explainable AI-based Decision Support for Nocturnal Hypoglycemia Prevention in Type 1 Diabetes
arXiv 2608.21563 · 2026-08-25Kruskal张量的边界秩等于秩，以及斜分解的Kruskal定理
Border rank=rank for Kruskal tensors and a Kruskal's theorem for skew decompositions
arXiv 2608.21562 · 2026-08-25学习轨迹上的损失参数化费希尔宽度
Loss-Parameterized Fisher Width Along Learning Trajectories
arXiv 2608.21561 · 2026-08-25受控退化下的证据状态可靠性：多阶段大语言模型（LLM）流水线中的解析器有效性偏差
Evidence-State Reliability Under Controlled Degradation: Parser-Validity Divergence in a Multi-Stage LLM Pipeline
arXiv 2608.21559 · 2026-08-25通过TRIAD实现多跳RAG评估自动化：从上下文提取到验证数据集生成
Automating Multi-Hop RAG Evaluation via TRIAD: From Context Extraction to Validated Dataset Generation
arXiv 2608.21558 · 2026-08-25异常值的影响：基于后果的检测
Outlier Impact: Detection by Consequences
arXiv 2608.21557 · 2026-08-25完全交数值半群的可实现（嵌入维数，重数）对的完整刻画
A Complete Characterization of Realizable (Embedding Dimension, Multiplicity) Pairs for Complete Intersection Numerical Semigroups
arXiv 2608.21556 · 2026-08-25张量寻求布局：为机器学习编译器形式化布局选择
Tensor Seeks Layout: Formalizing Layout Selection for ML Compilers
arXiv 2608.21555 · 2026-08-25分布偏移下异构多机器人任务分配的模型强化学习方法
Model-Based Reinforcement Learning for Heterogeneous Multi-Robot Task Assignment Under Distribution Shifts
arXiv 2608.21554 · 2026-08-25基于白光腔的超光速激光器的光注入锁定
Optical injection locking of white light cavity based superluminal lasers
arXiv 2608.21552 · 2026-08-25复域中的稀疏可分因子分析及其在局部场电位数据中的应用
Sparse Separable Factor Analysis in the Complex Domain with an Application to Local Field Potential Data
arXiv 2608.21551 · 2026-08-25GOLEM：面向电动汽车电池拆解的模块化人形自主系统
GOLEM: Modular Humanoid Autonomy Towards Electric Vehicle Battery Disassembly
arXiv 2608.21550 · 2026-08-25增强用户抵御AI增强型钓鱼攻击的能力：一种用于检测和个性化训练的两阶段框架
Enhancing User Resilience Against AI-Augmented Phishing: A Two-Stage Framework for Detection and Personalized Training
arXiv 2608.21547 · 2026-08-25非最小耦合引力下带电球对称各向异性物质构型坍缩的动力学不稳定性研究
Study of Dynamical Instability of Collapsing Charged Spherically Symmetric Anisotropic Matter Configurations within Non-Minimally Coupled Gravity
arXiv 2608.21546 · 2026-08-25一种基于时间可逆动态低秩近似的内存高效伴随状态优化方法
A Memory-Efficient Adjoint State Optimization Method Based on Time-Reversible Dynamical Low-Rank Approximation
arXiv 2608.21545 · 2026-08-25权重中遗忘，工具中恢复：面向大语言模型智能体的智能体工具遗忘
Forgotten in Weights, Recovered by Tools: Agentic Tool Unlearning for LLM Agents
arXiv 2608.21544 · 2026-08-25presto：基于想象搜索的高效无训练开放世界物体放置方法
presto: Efficient, Training-free, and Open-world Object Placement via Imaginary Search
arXiv 2608.21543 · 2026-08-25基于p轨道编码的锗空穴噪声鲁棒自旋轨道量子比特
Noise-Robust Spin-Orbit Qubit in Germanium Holes via p-Orbital Encoding
arXiv 2608.21542 · 2026-08-25超越稀疏权重：注意力机制何时可压缩？
Beyond Sparse Weights: When Is Attention Compressible?
arXiv 2608.21541 · 2026-08-25揭示晶界相微结构的自动化分析
Automated Analysis to Reveal Grain Boundary Phase Microstructures
arXiv 2608.21540 · 2026-08-25作为分布式漂移加惩罚控制问题的联邦持续学习
Federated Continual Learning as a Distributed Drift-Plus-Penalty Control Problem
arXiv 2608.21539 · 2026-08-25反德西特空间中的带电非阿贝尔黑弦
Electrically Charged Non-Abelian Black String in Anti-de Sitter Space
arXiv 2608.21538 · 2026-08-25面向自动招聘的反事实逐决策偏差审计：定位与解释申请人跟踪系统中的差异影响
Counterfactual, Per-Decision Bias Auditing for Automated Hiring: Localizing and Explaining Disparate Impact in Applicant Tracking Systems
arXiv 2608.21537 · 2026-08-25基于集成光子学的可编程传输中计算技术
Programmable Compute-in-Transit using Integrated Photonics
arXiv 2608.21536 · 2026-08-25亚循环p群的可分性、不可分性及自动可分性
Splittability, non-splittability, and automatic splittability of metacyclic $p$-groups
arXiv 2608.21535 · 2026-08-25正特征下广义Kummer曲面的分类
The classification of generalised Kummer surfaces in positive characteristic
arXiv 2608.21534 · 2026-08-25用于高效多机器人仓库作业的无模型自适应参数调优
Model-Free Adaptive Parameter Tuning for Efficient Multi-Robot Warehouse Operations
arXiv 2608.21533 · 2026-08-25手征对称性与黑洞等谱性
Chiral symmetry and black hole isospectrality
arXiv 2608.21532 · 2026-08-25用于需求工程的大语言模型：跨任务实证评估
Large Language Models for Requirements Engineering: A Cross-Task Empirical Evaluation
arXiv 2608.21531 · 2026-08-25基于加权集成学习的网球多模态损伤风险与表现预测
Multimodal Injury Risk and Performance Prediction in Tennis Using Weighted Ensemble Learning
arXiv 2608.21530 · 2026-08-25
