2026-10论文中文摘要
大初始化下逼近逻辑斯蒂梯度下降轨迹的理想路径
Ideal Paths for Approximating Logistic Gradient Descent Trajectories at Large Initialization
arXiv 2610.04142 · 2026-10-07图的正常帽子猜测数的一个上界
An upper bound on the proper hat guessing number of graphs
arXiv 2610.04124 · 2026-10-07结构化量子信号中全局极值的加速搜索
Accelerated search for global extrema in structured quantum signals
arXiv 2610.04024 · 2026-10-07VolS-GS：具有体积次表面散射的可重照明高斯溅射
VolS-GS: Relightable Gaussian Splatting with Volumetric Subsurface Scattering
arXiv 2610.04007 · 2026-10-07DePICT：面向约束下游任务的决策保持接口
DePICT: Decision-Preserving Interface for Constrained Downstream Tasks
arXiv 2610.03945 · 2026-10-07OncoNoteBERT：面向真实世界门诊肿瘤学笔记自然语言处理的基础表示模型
OncoNoteBERT: A Foundation Representation Model for Natural Language Processing of Real-World Outpatient Oncology Notes
arXiv 2610.03829 · 2026-10-07矩形网格图中的上支配数及其与子法定人数染色的关系
Upper $k$-Star-Forming Sets, $k$-Independence, and Upper Domination
arXiv 2610.03785 · 2026-10-07统计暗物质：协同作用无处不在，但难以捕捉
Statistical Dark Matter: Synergy is everywhere but is hard to capture
arXiv 2610.03764 · 2026-10-07奇特征下 $\mathbb F_q$ 上 MDS 猜想维度范围的线性 $q$ 区间
A linear-in-$q$ range of dimensions for the MDS conjecture over $\mathbb F_q$ in odd characteristic
arXiv 2610.03740 · 2026-10-07有限维量子理论的操作性刻画
An operational characterization of finite-dimensional quantum theory
arXiv 2610.03694 · 2026-10-07大色数的几何无三角形图
Geometric triangle-free graphs of large chromatic number
arXiv 2610.03517 · 2026-10-07任意过程的最小最大熵产生
Minimax entropy production for arbitrary processes
arXiv 2610.03478 · 2026-10-07功夫运动员机器人：从视频学习高动态人形运动并具备统一鲁棒恢复能力
KungfuAthleteBot: learning high-dynamic humanoid motion from video with unified robust recovery
arXiv 2610.03388 · 2026-10-07类型化决策模型中候选覆盖率的基准测试
Benchmarking candidate coverage and rejection policy transfer in typed decision models
arXiv 2610.03387 · 2026-10-07ALPS II 双激光外差探测系统
Two-Laser Heterodyne Detection System for ALPS II
arXiv 2610.03376 · 2026-10-07SCAD：面向长时程智能体的结构化信用分配与蒸馏
SCAD: Structured Credit Assignment and Distillation for Long-Horizon Agents
arXiv 2610.03372 · 2026-10-07量子码参数，可通过可证明大小的证书进行检验
Quantum error-correcting code parameters, checkable by a certificate of provable size
arXiv 2610.03214 · 2026-10-07倾斜三维节环半金属中的等离激元模式. II. 涡旋节环
Plasmon modes in tilted three-dimensional nodal-ring semimetals. II. Vortex nodal ring
arXiv 2610.03188 · 2026-10-07基于文本梯度的交易策略优化
Trading Strategy Optimization via Textual Gradient
arXiv 2610.03128 · 2026-10-07偏好的可验证、可表述与隐性组成部分
Verifiable, Articulable, and Tacit Components of Preference
arXiv 2610.03025 · 2026-10-07面向Schema-as-Code生物医学命名实体识别的指南增强多智能体框架
A Guideline-Augmented Multi-Agent Framework for Schema-as-Code Biomedical Named Entity Recognition
arXiv 2610.02970 · 2026-10-07LUMOS：从训练数据到行为输出追踪LLMs中的参数化知识
LUMOS: Tracing Parametric Knowledge from Training Data to Behavioral Outputs in LLMs
arXiv 2610.02902 · 2026-10-07从复杂分数量子霍尔边缘的散粒噪声中提取任意子电荷
Extracting the anyon charge from shot noise in complex fractional quantum Hall edges
arXiv 2610.02818 · 2026-10-07BitNest：用于内存高效LLM推理加速的位嵌套投机解码
BitNest: Bit-Nested Speculative Decoding for Memory-Efficient LLM Inference Acceleration
arXiv 2610.02800 · 2026-10-07PAPER2LLM++：从研究论文中持续自我进化的大语言模型
PAPER2LLM++: Continual Self-Evolution of LLMs from Research Papers
arXiv 2610.02793 · 2026-10-07将记忆与上下文解耦：面向令牌高效测试时持续学习的结构化记忆
Decoupling Memory from Context: Structured Memory for Token-Efficient Test-Time Continual Learning
arXiv 2610.02687 · 2026-10-07基于选择性状态空间模型的分布式学习：架构感知的收敛性分析
Distributed Learning with Selective State Space Models: Architecture-Aware Convergence Analysis
arXiv 2610.02659 · 2026-10-07OTA训练神经接收器中的跨站点迁移与空间遗忘
Cross-Site Transfer and Spatial Forgetting in OTA-Trained Neural Receivers
arXiv 2610.02577 · 2026-10-07大规模通用代数证明
Proving at Scale for Universal Algebra
arXiv 2610.02500 · 2026-10-07SOLO：基于仅扫描采样倒排列表的可认证召回率度量相似性搜索
SOLO: Certified-Recall Metric Similarity Search with Scan-Only Sampled Inverted Lists
arXiv 2610.02387 · 2026-10-07基于分叉的域分解方法与神经算子用于血流模拟
A Bifurcation-Based Domain Decomposition Method with Neural Operators for Blood Flow Simulation
arXiv 2610.02238 · 2026-10-07一种面向高压缩比、高吞吐量的自适应异构无损压缩架构
An Adaptive Heterogeneous Architecture for High-Ratio, High-Throughput Lossless Compression
arXiv 2610.02236 · 2026-10-07单关系群的 Sofic 性
Soficity of One-Relator Groups and Reducible Presentations
arXiv 2610.02213 · 2026-10-07有限记忆量子过程的序贯容量
Sequential Capacity of Quantum Processes with Finite Memory
arXiv 2610.02068 · 2026-10-07DiDE：用于3D风格化的颜色-纹理解耦直接注入
DiDE:Direct Injection with Color-Texture DEcoupling for 3D Stylization
arXiv 2610.02044 · 2026-10-07局部维度 $\geq 3$ 的均场交换模型的亚稳定性
On the Metastability of the Mean-Field Interchange Model for Local Dimension $\geq 3$
arXiv 2610.01987 · 2026-10-07绘制RAG领域图谱：效率、防御、交互与推理的四轴分类法
Rethinking Knowledge Retrieval for Generation: A Survey on RAG Architectures and Applications
arXiv 2610.01936 · 2026-10-07正则化最优输运的全局正则性与稀疏性
Global regularity and sparsity for regularised optimal transport
arXiv 2610.01928 · 2026-10-07SkillEvoLean：面向 Lean 证明器的突变增强技能进化
SkillEvoLean: Mutation-enhanced skill evolution for Lean provers
arXiv 2610.01799 · 2026-10-07Ivanov 关于 Torelli 群子群的有限指数 Frattini 子群猜想
Ivanov's finite-index Frattini conjecture for subgroups of mapping class groups
arXiv 2610.01734 · 2026-10-07面向世界行动模型的完成感知引导
Completion Aware Guidance for World Action Models
arXiv 2610.01559 · 2026-10-07Connes嵌入猜想的一个显式多项式反例
An Explicit Polynomial Counterexample to Connes' Embedding Conjecture
arXiv 2610.01536 · 2026-10-07第二个模型何时有帮助？大语言模型验证中的跨模型评审
When Does a Second Model Help? Cross-Model Review in LLM Verification
arXiv 2610.01471 · 2026-10-07SpikeMoE：受大脑启发的竞争路由用于灵活脉冲混合专家模型
SpikeMoE: Brain-Inspired Competitive Routing for Flexible Spiking Mixture-of-Experts
arXiv 2610.01418 · 2026-10-07第三Borwein猜想的一个证明
A Proof of the Third and Cubic Borwein Conjectures
arXiv 2610.01156 · 2026-10-07ReSolve：通过选择性生成式调节重用候选推理
ReSolve: Reusing Candidate Reasoning through Selective Generative Moderation
arXiv 2610.01140 · 2026-10-07扩展强化学习真的需要更多训练吗？
Does Scaling Reinforcement Learning Really Require More Training?
arXiv 2610.01133 · 2026-10-07语言模型能从测试时计算中获得多少收益？
How Much Can Language Models Gain from Test-Time Computation?
arXiv 2610.01110 · 2026-10-07AgSpec：在编码智能体流程中突破基于检索的推测解码的极限
AgSpec: Pushing the Limits of Retrieval-Based Speculative Decoding in Coding Agent Pipelines
arXiv 2610.01108 · 2026-10-07对抗性线性约束马尔可夫决策过程的最优速率算法
Rate-Optimal Algorithm for Adversarial Linear CMDPs
arXiv 2610.00927 · 2026-10-07
