2026-10论文中文摘要
灵活刚性：将通用可信制备协议编译为顺序设备无关协议
Flexible rigidity: Compiling generic trusted-preparation protocols into sequential device-independent protocols
arXiv 2610.00444 · 2026-10-02鲁棒的经典与量子通信在几乎独立同分布信道上的实现
Robust classical and quantum communication over almost-iid channels
arXiv 2610.00442 · 2026-10-02量子秘密共享与纠错 vs 不可克隆定理
Quantum Secret Sharing and Error Correction vs No-Cloning
arXiv 2610.00441 · 2026-10-02有/无经典反馈的量子擦除信道上的有限块长经典通信
Finite-blocklength classical communication over the quantum erasure channel with and without classical feedback
arXiv 2610.00440 · 2026-10-02三味中微子振荡的无信号时间条件
No-signaling-in-time condition for three-flavor neutrino oscillations
arXiv 2610.00439 · 2026-10-02迈向通用人形移动操作模型：基于以自我为中心全身人类数据预训练
Towards a General Humanoid Loco-Manipulation Model via Egocentric Whole-Body Human Data Pretraining
arXiv 2610.00438 · 2026-10-02JevSpawn：通过组合动作空间实现自适应智能体推理
JevSpawn: Adaptive Agentic Inference through Compositional Action Spaces
arXiv 2610.00437 · 2026-10-02每个批次都是其自身的验证集：用于大语言模型微调中在线数据选择的留一法梯度匹配
Every Batch Is Its Own Validation Set: Leave-One-Out Gradient Matching for Online Data Selection in LLM Fine-Tuning
arXiv 2610.00436 · 2026-10-02AI智能体如何发现科学公式：从Hydrotope再发现到新的水波振幅
How AI Agents Discover Scientific Equations: From Hydrotope Rediscovery to New Water-Wave Amplitudes
arXiv 2610.00435 · 2026-10-02Smyth在正四面体中的自由边界极小圆盘的精确超几何Weierstrass表示
An exact hypergeometric Weierstrass representation for Smyth's free-boundary minimal disk in the regular tetrahedron
arXiv 2610.00434 · 2026-10-02立方根零代数的Big Finitistic维数
Big Finitistic Dimensions of Radical-Cube-Zero Algebras
arXiv 2610.00433 · 2026-10-02XOR-Trellis：超低复杂度反量化与曲率感知的无Hadamard大语言模型量化
XOR-Trellis: Ultra-Low-Complexity Dequantization and Curvature-Aware Hadamard-Free LLM Quantization
arXiv 2610.00432 · 2026-10-02ChainLoRA：用于大语言模型持续学习的几何保持任务向量合并
ChainLoRA: Geometry-Preserving Task Vector Merging for Continual Learning in LLMs
arXiv 2610.00431 · 2026-10-02间歇性隔离在单纯减少接触之外抑制流行病传播
Intermittent quarantine suppresses epidemic spreading beyond simple contact reduction
arXiv 2610.00429 · 2026-10-02活性物质电池
Active-Matter Battery
arXiv 2610.00428 · 2026-10-02调制增强与星座整形用于LDPC编码QPSK
Modulation Augmentation and Constellation Shaping for LDPC-Coded QPSK
arXiv 2610.00427 · 2026-10-02IrekoGPT：将结构化剪枝转化为事后可瘦身的大语言模型
IrekoGPT: Turning Structured Pruning into Post-Hoc Slimmable LLMs
arXiv 2610.00426 · 2026-10-02代码能运行，环境却不行：衡量AI生成软件的环境可复现性
Code That Works, Environments That Don't: Measuring Environment Reproducibility in AI-Generated Software
arXiv 2610.00425 · 2026-10-02因果修复的目标依赖极限：高斯模型中的前导对数前沿
Target-Dependent Limits of Causal Repair: A Leading-Log Frontier in a Gaussian Model
arXiv 2610.00424 · 2026-10-02学习局部覆盖：硬信息视野下的图神经组合优化
Learning to Cover Locally: Graph Neural Combinatorial Optimization under a Hard Information Horizon
arXiv 2610.00422 · 2026-10-02持久的分数，泄漏的基准：衡量公开脑肿瘤MRI分类中的数据集污染
Scores That Hold, Benchmarks That Leak: Measuring Dataset Contamination in Public Brain-Tumor MRI Classification
arXiv 2610.00421 · 2026-10-02可迁移的图元网络
Transferable Graph Metanetworks
arXiv 2610.00420 · 2026-10-02ProxyMOS：通过多教师蒸馏与自适应路由实现无标签语音质量评估
ProxyMOS: Label-Free Speech Quality Assessment by Multi-Teacher Distillation with Adaptive Routing
arXiv 2610.00419 · 2026-10-02CommunityKV：通过图分区实现高效长上下文解码
CommunityKV: Efficient Long-Context Decoding via Graph Partitioning
arXiv 2610.00418 · 2026-10-02来源识别不是适应性测试：衡量合成数据归因的极限
Source Identification Is Not Fitness Testing: Measuring the Limits of Synthetic-Data Attribution
arXiv 2610.00417 · 2026-10-02用国际象棋基准测试大语言模型的提示优化
Benchmarking Prompt Optimization of Large Language Models With Chess
arXiv 2610.00416 · 2026-10-02更好的分数意味着更好的物理吗？Sim2Real 神经算子的物理基础解释
Do Better Scores Mean Better Physics? Physics-Grounded Explanations for Sim2Real Neural Operators
arXiv 2610.00415 · 2026-10-02从图像潜空间到模糊规则：胃肠基础模型的可解释分析
From Image Latent Space to Fuzzy Rules: Interpretable Analysis of Gastrointestinal Foundation Model
arXiv 2610.00414 · 2026-10-02模范畴的 magnitude 与特殊双列代数
Magnitude of module categories and special biserial algebras
arXiv 2610.00413 · 2026-10-02EchoPress：通过虚拟上下文重建实现查询无关的KV缓存剪枝
EchoPress: Query-Agnostic KV Cache Pruning via Virtual Context Reconstruction
arXiv 2610.00412 · 2026-10-02VANDAM：利用DNA分子先验查看核苷酸序列
VANDAM: Viewing a nucleotide sequence with DNA molecular priors
arXiv 2610.00411 · 2026-10-02弯曲时空中的费米子粒子产生：物理布洛赫矢量动力学与精确德西特标度
Fermionic Particle Creation in Curved Spacetime: Physical Bloch-Vector Dynamics and Exact de Sitter Scaling
arXiv 2610.00410 · 2026-10-02编织曲面与小 exotic 4-流形
Braided Surfaces and Small Exotic 4-Manifolds
arXiv 2610.00409 · 2026-10-02UniBuc在SemEval-2024任务2中：使用Solar进行针对临床自然语言推理的定制提示
UniBuc at SemEval-2024 Task 2: Tailored Prompting with Solar for Clinical NLI
arXiv 2610.00408 · 2026-10-02规范SLE可逆的Weil--Petersson响应扩散
The canonical SLE-reversible Weil--Petersson response diffusion
arXiv 2610.00407 · 2026-10-02LLM作为低资源语言评判者：为罗马尼亚语适配Ragas与比较排序
LLM-as-a-Judge for Low-Resource Languages: Adapting Ragas and Comparative Ranking for Romanian
arXiv 2610.00406 · 2026-10-02RACE：面向邻居丰富时间序列基础模型预测的残差感知测试时自适应
RACE: Residual-Aware Test-Time Adaptation for Neighbor-Rich Time-Series Foundation Model Forecasting
arXiv 2610.00405 · 2026-10-02基于部分视觉观测的全身空中抓取与提升
Whole-Body Aerial Grasping and Lifting via Partial Visual Observations
arXiv 2610.00404 · 2026-10-02不和谐的芭蕾舞演员与狡猾的胡萝卜：意大利脑腐内容的比较多模态分析
Dissonant ballerinas and crafty carrots: a comparative multi-modal analysis of Italian brain rot
arXiv 2610.00402 · 2026-10-02从旧到新的 Starobinsky 暴胀
From the Old to the New Starobinsky Inflation
arXiv 2610.00401 · 2026-10-02表示转变揭示多轮LLM智能体中新兴的安全风险
Representation Transitions Reveal Emerging Safety Risks in Multi-Turn LLM Agents
arXiv 2610.00400 · 2026-10-02基于实例学习理论的能量模型中的元认知推理
Metacognitive Reasoning in Energy Based Models using Instance Based Learning Theory
arXiv 2610.00399 · 2026-10-02WIPSNet：基于夜间阻抗呼吸描记术的儿科喘息检测深度学习
WIPSNet: Deep Learning for Paediatric Wheeze Detection from Overnight Impedance Pneumography
arXiv 2610.00398 · 2026-10-02NEUROTOKEN：基于条件流匹配的包络解码联合源与方向听觉注意解码
NEUROTOKEN: Joint Source and Directional AAD with Envelope Decoding via Conditional Flow Matching
arXiv 2610.00397 · 2026-10-02广义Charlier递推系数的完全阶乘渐近与可积结构
Complete Factorial Asymptotics and Integrable Structure of Generalized Charlier Recurrence Coefficients
arXiv 2610.00396 · 2026-10-02特异性感知扩散引导：基于方差缩减的顺序蒙特卡洛方法
Specificity-Aware Diffusion Steering via Variance-Reduced Sequential Monte Carlo
arXiv 2610.00395 · 2026-10-02分叉：重放下的突然过拟合
Forking: Sudden Overfitting Under Replay
arXiv 2610.00394 · 2026-10-02修正引力中Reissner-Nordström黑洞的准正则模
Quasinormal Modes of Reissner-Nordström Black Holes in Modified Gravity
arXiv 2610.00393 · 2026-10-02从A2A攻击到包络层防御：LLM智能体的红队评估与三层同构攻防模型
From A2A Attacks to Envelope-Layer Defense: Red-Teaming Evaluation of LLM Agents and a Three-Layer Isomorphic Attack-Defense Model
arXiv 2610.00392 · 2026-10-02使用模糊认知图的可解释合成医学表格数据生成，用于临床决策支持
Interpretable Synthetic Medical Tabular Data Generation for Clinical Decision Support Using Fuzzy Cognitive Maps
arXiv 2610.00391 · 2026-10-02
