2026-09论文中文摘要
从原始可解性复杂度指标证明到Weihrauch度
From raw Solvability Complexity Index proofs to Weihrauch degrees
arXiv 2609.20378 · 2026-09-21RoboFind：面向盲人或低视力人群的多智能体个性化物体搜索
RoboFind: Multi-Agent Personalized Object Search for People Who Are Blind or Have Low Vision
arXiv 2609.20330 · 2026-09-21NeuSOGA3D：一种用于可解释三维几何重建的神经符号框架
NeuSOGA3D: A Neuro-Symbolic Framework for Explainable 3D Geometric Reconstruction
arXiv 2609.20323 · 2026-09-21高秩图的Kumjian--Pask代数的诱导表示与限制表示
Maximal tails,character fibres and induced modules for pullback Kumjian-Pask algebras
arXiv 2609.20230 · 2026-09-21AR(1) Karhunen-Loeve变换的精确快速分解
Exact fast factorizations of the AR(1) Karhunen-Loeve transform
arXiv 2609.20221 · 2026-09-21小到足以知晓一切：完全可枚举Transformer作为延迟泛化科学的仪器
Small Enough to Know Everything: The Fully-Enumerable Transformer as an Instrument for the Science of Delayed Generalization
arXiv 2609.20166 · 2026-09-21缺血性卒中分割及多中心非增强CT上净水摄取量化的监督目标域自适应
Ischemic Stroke Segmentation and Net Water Uptake Quantification on Multicenter Non-Contrast CT Using Supervised Target-Domain Adaptation
arXiv 2609.20151 · 2026-09-21量子首次通过统计中的例外点与Jordan链特征
An exact Jordan signature in cumulative-count first-passage statistics
arXiv 2609.20148 · 2026-09-21GPT-6-Astra 在导航工作流中的应用：连续环境中零样本视觉与语言导航的行为分析
How Far Can GPT-6-Astra Go? Evaluating Capabilities in Zero-Shot Vision-and-Language Navigation
arXiv 2609.20116 · 2026-09-21AnyviewMeter：利用相机几何与多视角注意力适配机器人奖励模型
AnyviewMeter: Adapting Robotic Reward Models with Camera Geometry and Multi-View Attention
arXiv 2609.20106 · 2026-09-21安全关键场景从初始场景中涌现
Safety-Critical Scenarios Emerge from Initial Scenes
arXiv 2609.20103 · 2026-09-21估计网络中的噪声调整换手率
Noise-adjusted turnover in estimated networks
arXiv 2609.20044 · 2026-09-21Hopper：面向拜占庭容错对等采样的有界内存协作去偏
Hopper: Bounded-Memory Collaborative Debiasing for Byzantine-Tolerant Peer Sampling
arXiv 2609.19893 · 2026-09-21PACE：精确AI电影化表达——面向脚本预可视化与几何一致性的类型化规范
PACE: Precise AI Cinematic Expression
arXiv 2609.19853 · 2026-09-21共识引导的共享-特定三视图学习用于语音情感识别
Consensus-Guided Shared-Specific Tri-View Learning for Speech Emotion Recognition
arXiv 2609.19826 · 2026-09-21CoRELoop：用于音频深度伪造检测的参数高效受控循环精化
CoReLoop: Parameter-Efficient Controlled Recurrent Refinement for Audio Deepfake Detection
arXiv 2609.19818 · 2026-09-21LIFD：用于机器人操作中三维感知场景记忆的锚定扩散
LIFD: Anchored Diffusion for 3D-Aware Scene Memory in Robotic Manipulation
arXiv 2609.19796 · 2026-09-21转导实现的超导量子中继器：面向高保真门下的确定性纠缠分发
Transduction-Enabled Superconducting Quantum Repeater: Toward Deterministic Entanglement Distribution with High-Fidelity Gates
arXiv 2609.19784 · 2026-09-21重新思考多智能体协作：何时多即是少
Rethinking Multi-Agent Collaboration: When More Is Less
arXiv 2609.19759 · 2026-09-21伸展树顺序访问定理的精细化分析
A Refined Analysis of the Sequential Access Theorem for Splay Trees
arXiv 2609.19746 · 2026-09-21筛选出贫困者：SNAP工作要求的影响
Screening Out the Needy: The Effects of SNAP Work Requirements
arXiv 2609.19660 · 2026-09-21量子计算机不会那么不同：面向规模的量子计算机架构蓝图
Quantum computers will not be that different: A blueprint for quantum computer architecture at scale
arXiv 2609.19639 · 2026-09-21OmniCalib：人形机器人的无靶标、任务结构化自标定
OmniCalib: Target-Free, Task-Structured Self-Calibration for Humanoid Robots
arXiv 2609.19582 · 2026-09-21大型语言模型智能体用于基于证据的遗传疾病严重程度分类
Large Language Model Agents for Evidence Based Genetic Disease Severity Classification
arXiv 2609.19569 · 2026-09-21PerSeM：用于长时程开放词汇无人机测绘的持久语义记忆
PerSeM: Persistent Semantic Memory for Long-Horizon Open-Vocabulary UAV Mapping
arXiv 2609.19542 · 2026-09-21可信大型语言模型、智能体AI与多模态系统的统一评估框架
A Unified Evaluation Framework for Trustworthy Large Language Models, Agentic AI, and Multimodal Systems
arXiv 2609.19524 · 2026-09-21用CLAS12探测器测量$\gamma_v p \to \pi^+ \pi^-p'$在$Q^2$为2.4-8.0 GeV$^2$和$W$为1.4-2.1 GeV范围内的截面
Measurements of $γ_v p \to π^+ π^-p'$ Cross Sections with the CLAS12 Detector for $Q^2$ from 2.4-8.0 GeV$^2$ and $W$ from 1.4-2.1 GeV
arXiv 2609.19395 · 2026-09-21尖点系数的严格SDE比较与反例
Strict SDE Comparison for Cusp Coefficients and Counterexamples
arXiv 2609.19389 · 2026-09-21细粒度危害信号在LLM安全中的作用
The Role of Fine-grained Harm Signals in LLM Safety
arXiv 2609.19366 · 2026-09-21难解枚举问题如同俄罗斯套娃：二维二次格子上单体-二聚体覆盖的结构性质
Intractable enumeration problems are like Russian nesting dolls: structural properties of monomer-dimer coverings on two-dimensional quadratic lattices
arXiv 2609.19231 · 2026-09-21Epic：存储内计算的高效编程范式
Programming In-Storage Computing with Located, Stateful Dataflow
arXiv 2609.19206 · 2026-09-21采样揭示风格：无监督、免训练发现LLM激活中的提示条件风格轴
Sampling Reveals Style: Unsupervised, Training-Free Discovery of Prompt-Conditional Stylistic Axes in LLM Activations
arXiv 2609.19150 · 2026-09-21空芯光纤系统中的逐信道发射功率优化：利用解耦的功率预算
Per-Channel Launch-Power Optimization in Hollow-Core Fiber Systems
arXiv 2609.18977 · 2026-09-21ASLEval：衡量LLM智能体会话中的隐私暴露位移
ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions
arXiv 2609.18864 · 2026-09-21各向异性极小超流边界点处平坦爆破的存在性
Existence of flat blowups at boundary points of anisotropic minimizing hypercurrents
arXiv 2609.18838 · 2026-09-21最小成本生成树博弈的核心稳定性识别：参数化视角
Core stability recognition for minimum-cost spanning tree games: Parameterized perspective
arXiv 2609.18807 · 2026-09-21量化HCLs对固定微架构MXFP4加速器的影响
Quantifying the Effect of HCLs on a Fixed-Microarchitecture MXFP4 Accelerator
arXiv 2609.18792 · 2026-09-21体积泛函临界度量的谱性质
Spectral properties for critical metrics of the volume functional
arXiv 2609.18717 · 2026-09-21迈向可组合的网络数字孪生：基于子图的延迟预测研究
Toward Composable Network Digital Twins: A Subgraph-Based Latency Prediction Study
arXiv 2609.18704 · 2026-09-21$\mathbb{R}^N$ 与紧致度量图乘积上的非线性薛定谔方程
The nonlinear Schrödinger equation on products of $\mathbb{R}^N$ and compact metric graphs
arXiv 2609.18698 · 2026-09-21有色噪声谱图的零点
Zeros of the Spectrogram of Colored Noise
arXiv 2609.18689 · 2026-09-21通过潜在神经符号推理解缠长期记忆
Disentangling Long-Term Memory via Latent Neuro-Symbolic Reasoning
arXiv 2609.18461 · 2026-09-21桥接磁热风与光蒸发以模拟盘消散
Bridging magnetothermal winds and photoevaporation to model discs dispersal
arXiv 2609.18334 · 2026-09-21基于知识图谱的增强与检索增强生成在文化相关问答中的对比
Knowledge-Graph Based Augmentation versus Retrieval Augmented Generation for Cultural-Related Question Answering
arXiv 2609.18317 · 2026-09-21SEA-LION-v4.8：技术报告
SEA-LION-v4.8: A Technical Report
arXiv 2609.18310 · 2026-09-21CPR：在连续自回归建模中结合全局作曲、局部演奏与全序列精修的钢琴渲染
CPR: Combining global composing, local performing and full-sequence refining in piano rendering with continuous autoregressive modelling
arXiv 2609.18216 · 2026-09-21PentestChain：一种成本感知的、基于MCP编排的自动化渗透测试框架，使用免费层LLM
PentestChain: A Cost-Aware, MCP-Orchestrated Framework for Automated Penetration Testing with Free-Tier LLMs
arXiv 2609.18120 · 2026-09-21可测动力学中的有界链式性质
Bounded chaining in measurable dynamics
arXiv 2609.18061 · 2026-09-21复杂躯干姿态的几何捷径：双螺旋耦合实现低维控制
Geometric Shortcuts for Complex Trunk Postures: Dual-Helicity Coupling Enables Low-Dimensional Control
arXiv 2609.18050 · 2026-09-21VeriBugBench：一个基于经验构建Verilog RTL调试基准的框架
VeriBugBench: An Empirically Grounded Framework for Constructing Verilog RTL Debugging Benchmarks
arXiv 2609.18022 · 2026-09-21
