2026-10论文中文摘要
EgoAlign：弥合人形机器人与人之间的差距以实现远程操作-操控
EgoAlign: Bridging the Human-Humanoid Gap for Long-Range Loco-Manipulation
arXiv 2609.38046 · 2026-10-05可审计的长期记忆：确定性检索链在LongMemEval-S上测得479/475（满分500）
Auditing Long-Term Memory Evaluation: Repeated Judging, Reader Variation, and Negative Controls
arXiv 2609.38021 · 2026-10-05Lazard 关于 N-序列的实现问题：Bar 障碍与有限商下降
Lazard's Realization Problem for N-Series: Bar Obstructions and Finite-Quotient Descent
arXiv 2609.38002 · 2026-10-05拜占庭因果可靠广播：恒定元数据开销
Byzantine Causal Reliable Broadcast with Constant Metadata Overhead
arXiv 2609.37913 · 2026-10-05权重读取与写入特征：基于激活空间的可扩展参数分解
Weights Read and Write Features: Scalable Parameter Decomposition Grounded in Activation Space
arXiv 2609.37731 · 2026-10-05WISE-ATTA：在预算受限的主动测试时适应中何时请求标签
WISE-ATTA: When to Ask for Labels in Budgeted Active Test-Time Adaptation
arXiv 2609.37687 · 2026-10-05TomoTransformer：面向CT重建的基础模型
TomoTransformer: Towards a Foundation Model for CT Reconstruction
arXiv 2609.37605 · 2026-10-05图条件化的在线策略智能体蒸馏：从现成教师模型学习
Graph-Conditioned On-Policy Agent Distillation from Off-the-Shelf Teachers
arXiv 2609.37522 · 2026-10-05非交换电路的有理恒等式检验属于多项式空间
Rational Identity Testing for Noncommutative Circuits is in Polynomial Space
arXiv 2609.37319 · 2026-10-05从学习者行为到可复用技能：实现高效且有效的学习者模拟
From Learner Behavior to Reusable Skills for Effective and Efficient Learner Simulation
arXiv 2609.37157 · 2026-10-05ARGOS：计算连续体中服务编排的强化学习驱动多维弹性
ARGOS: Reinforcement Learning-Driven Multidimensional Elasticity for Service Orchestration in the Computing Continuum
arXiv 2609.37085 · 2026-10-05嵌入式双时相建筑损毁评估用于星上数据缩减
Embedded Bi-Temporal Building Damage Assessment for On-Board Data Reduction
arXiv 2609.37013 · 2026-10-05XBDD：一种带有逐边变量翻转映射的高度优化的ROBDD
XBDD: A Highly Optimized ROBDD with Per-Edge Variable-Flip Maps
arXiv 2609.36778 · 2026-10-05SafeCoEvo：测试时协同进化LLM智能体的安全护栏与守卫
SafeCoEvo: Co-Evolving Safety Harnesses and Guards for LLM Agents at Test-Time
arXiv 2609.36580 · 2026-10-05SCOPE：稀疏PDE推断的观测条件全目标预测
SCOPE: Observation-Conditioned Full-Target Prediction for Sparse PDE Inference
arXiv 2609.36527 · 2026-10-05强 EFX 可定向图的 polynomial time 刻画
A Polynomial Time Characterization For Strongly EFX Orientable Graphs
arXiv 2609.36498 · 2026-10-05Copula 活动子空间 I：用于降阶非高斯密度估计的得分-协方差方法
Copula Active Subspaces I: A Score-Covariance Method for Reduced-Order Non-Gaussian Density Estimation
arXiv 2609.36142 · 2026-10-05为什么对神经网络进行后门攻击如此容易？
Why Backdooring Neural Networks is so Easy?
arXiv 2609.36117 · 2026-10-05随机凸优化的无梯度方法及其随机函数约束
Gradient-Free Methods for Stochastic Convex Optimization with Stochastic Functional Constraints
arXiv 2609.35946 · 2026-10-05HERO：用于肿瘤学稳健表征的组织学编码器
HERO: Histology Encoder for Robust Representation in Oncology
arXiv 2609.35943 · 2026-10-05Ising介子的离散标度不变性
Discrete Scale Invariance of Ising Mesons
arXiv 2609.35940 · 2026-10-05为临床做好准备？临床试验中响应自适应随机化的开源软件调查
Ready for the Clinic? A Survey of Open-Source Software for Response-Adaptive Randomization in Clinical Trials
arXiv 2609.35939 · 2026-10-05CipherGenome：基因组混合专家模型的同态推理
CipherGenome: Homomorphic Inference for Genomic Mixture-of-Experts
arXiv 2609.35883 · 2026-10-05GenomeOcean Anywhere：基因组MoE的私有WebGPU推理
GenomeOcean Anywhere: Private WebGPU Inference for Genome MoEs
arXiv 2609.35882 · 2026-10-05GenoTrace：基因组基础模型蒸馏的可继承水印
GenoTrace: Inheritable Watermarks for Genome Foundation Model Distillation
arXiv 2609.35881 · 2026-10-05从 Pass@K 与 Pass@1 之间的差距中学习
Learning from the Gap Between Pass@K and Pass@1
arXiv 2609.35793 · 2026-10-05统一一步式视觉生成的分布训练
Unifying Distributional Training for One-Step Visual Generation
arXiv 2609.35763 · 2026-10-05正则化的可证明优势：对抗模仿学习的快速收敛率
Provable Benefits of Regularization: Fast Rates for Adversarial Imitation Learning
arXiv 2609.35698 · 2026-10-05从迁移到校准：跨模型、司法管辖区和规模保持智能体能力
From Migration to Calibration: Preserving Agent Capabilities across Models, Jurisdictions, and Scale
arXiv 2609.35149 · 2026-10-05自动化特征工程、AutoML 与决策聚焦学习用于改进能源消耗预测
Automated Feature Engineering, AutoML, and Decision-Focused Learning for Improved Energy Consumption Forecasting
arXiv 2609.35013 · 2026-10-05对数非线性电动力学中的视界、零轨道与光子球相结构
Horizon, Null-Orbit, and Photon-Sphere Phase Structures in Logarithmic Nonlinear Electrodynamics
arXiv 2609.34957 · 2026-10-05变换对齐的学习特征用于有损点云属性压缩
Transform-Aligned Learned Features for Lossy Point Cloud Attribute Compression
arXiv 2609.34834 · 2026-10-05高维因果外推的风险校准平衡
Risk-Calibrated Balancing for High-Dimensional Causal Extrapolation
arXiv 2609.34482 · 2026-10-05MASCIT：面向自然不规则时间序列的掩码感知状态空间分类器
MASCIT: A Mask-Aware State Space Classifier for Naturally Irregular Time Series
arXiv 2609.34409 · 2026-10-05WB-WAM：用于人形机器人移动操作任务的异构身体-手部预训练
WB-WAM: Heterogeneous Body-Hand Pre-training for Humanoid Loco-Manipulation
arXiv 2609.34199 · 2026-10-05理解使用大型语言模型的临床认知对话
Understanding Clinical Cognitive Dialogues Using Large Language Models
arXiv 2609.34125 · 2026-10-05关于最小化凸Hölder光滑函数的Active-Hull方法的最优线性最小化预言复杂度
On the Optimal Linear Minimization Oracle Complexity of Active-Hull Methods for Minimizing Convex Hölder Smooth Functions
arXiv 2609.34101 · 2026-10-05MaskCoFT：面向内存高效MoE推理的掩码协同自适应微调
MaskCoFT: Masked Co-Adaptive Fine-Tuning for Memory-Efficient MoE Inference
arXiv 2609.34077 · 2026-10-05在变化的世界中学习合作：对未来关注如何促进跨尺度合作
Learning to cooperate in a changing world: How caring about the future promotes cooperation across scales
arXiv 2609.34005 · 2026-10-05多任务职业生涯轨迹的凹过程：评估NBA表现的多项指标中的老化效应
Concave Processes for Multi-Task Career Trajectories: Evaluating Aging Across Multiple Measures of NBA Performance
arXiv 2609.33752 · 2026-10-05Bernstein定理的短证明：基于Omori-Yau最大值原理
A short proof of Bernstein's theorem via the Omori--Yau maximum principle
arXiv 2609.33643 · 2026-10-05线排列的广义 Terao 猜想
A generalized Terao Conjecture for line arrangements
arXiv 2609.33584 · 2026-10-05不透明指针下的神经符号间接调用分析
Neuro-Symbolic Indirect-Call Analysis under Opaque Pointers
arXiv 2609.33547 · 2026-10-05广义Ceresa环的Beilinson--Bloch高度
The Beilinson--Bloch heights of generalized Ceresa cycles
arXiv 2609.33223 · 2026-10-05CoLMbo-SV：一种用于可解释说话人验证的接地语言模型
CoLMbo-SV: A Grounded Language Model for Explainable Speaker Verification
arXiv 2609.33212 · 2026-10-05TimelyDAgger：面向VLA策略改进的时序感知专家查询
TimelyDAgger: Timing-Aware Expert Querying for VLA Policy Improvement
arXiv 2609.33157 · 2026-10-05解锁测地线Gromov-Wasserstein距离用于3D建模
Unlocking Geodesic Gromov-Wasserstein Distances for 3D Modeling
arXiv 2609.32824 · 2026-10-05$T^5$：强化中间训练中用于令牌级思维的双评论家训练
$T^5$: Twin-Critic Training for Token-Level Thoughts in Reinforcement Mid-Training
arXiv 2609.32791 · 2026-10-05用 $d-2$ 个额外顶点完成壳化
Completing shellings with $d-2$ extra vertices
arXiv 2609.32721 · 2026-10-05ProcGen 泛化差距衡量的是什么？动作规则、收敛性与缺失的随机基线
What Does a ProcGen Generalization Gap Measure? Action Rules, Residual Entropy, and the Missing Random Floor
arXiv 2609.32532 · 2026-10-05
