2026-10论文中文摘要
Xenia：用于天体物理多流体动力学的表驱动代码
Xenia: A Table-Driven Code for Astrophysical Multi-Fluid Hydrodynamics
arXiv 2610.04655 · 2026-10-06确定锕系元素中j特异性的未占据5f电子结构
Determining the j-Specific Unoccupied 5f Electronic Structure in Actinides
arXiv 2610.04654 · 2026-10-06高次谐波产生中利用等离子体重塑驱动场的相位匹配
Phase-matching in high-order harmonic generation with a plasma-reshaped driving field
arXiv 2610.04653 · 2026-10-06基于原始-对偶正则化的帕累托改进对抗攻击
Pareto-Improving Adversarial Attacks with Primal-Dual Regularization
arXiv 2610.04652 · 2026-10-06GS-Codec：用于神经音频编码的高斯溅射瓶颈
GS-Codec: A Gaussian-Splatting Bottleneck for Neural Audio Coding
arXiv 2610.04651 · 2026-10-06经典开普勒-库仑问题的动力学拟对称性与拉普拉斯-龙格-楞次矢量的守恒
Dynamical quasi-symmetry of the classical Kepler-Coulomb problem and the conservation of the Laplace-Runge-Lenz vector
arXiv 2610.04650 · 2026-10-06门控目标传播用于持续学习中的组合泛化
Gated Target Propagation for Compositional Generalization in Continual Learning
arXiv 2610.04649 · 2026-10-06限制模拟量子关联的双向平均通信成本
Bounding Two-Way Average Communication Cost of Simulating Quantum Correlations
arXiv 2610.04648 · 2026-10-06星形胶质细胞反馈和噪声对呼吸神经簇的影响
Influence of astrocyte feedback and noise on breathing neural clusters
arXiv 2610.04647 · 2026-10-06SEIS：自进化推理系统
SEIS: Self-Evolving Inference Systems
arXiv 2610.04646 · 2026-10-06mad 猜想的证明及其染色应用
Proof of the mad conjecture and its coloring applications
arXiv 2610.04645 · 2026-10-06重新审视神经图编辑距离模型的泛化能力
Revisiting the Generalization of Neural Graph Edit Distance Models
arXiv 2610.04644 · 2026-10-06大脑中的异archy：控制是一个谱系，而非指挥链
Heterarchy in the Brain: Control as a Spectrum, Not a Chain of Command
arXiv 2610.04643 · 2026-10-06Grounding Probes：从观察者模型隐藏状态进行与生成器无关的幻觉检测
Grounding Probes: Generator-Independent Hallucination Detection from Observer Model Hidden States
arXiv 2610.04642 · 2026-10-06回到未来：从LLM回归可读性特征
Back to the Future: Regressing Readability Features from LLMs
arXiv 2610.04641 · 2026-10-06超图表示学习中的超边随机效应
Hypergraph Representation Learning with Hyperlink Random Effects
arXiv 2610.04640 · 2026-10-06当谈话不是代码：比较模拟与开发软件的LLM智能体
When Talk Isn't Code: Comparing LLM Agents That Simulate and Develop Software
arXiv 2610.04639 · 2026-10-06策略即数据：基于优势回归的重放策略对偶平均
Policy as Data: Replay-Based Policy Dual Averaging via Advantage Regression
arXiv 2610.04638 · 2026-10-06多变量Herglotz-Nevanlinna函数的算子表示
Operator representations of Herglotz-Nevanlinna functions in several variables
arXiv 2610.04637 · 2026-10-06神经复杂度是否提升健康错误信息检测？一个泄漏控制的跨语料基准
Does Neural Complexity Improve Health Misinformation Detection? A Leakage-Controlled Cross-Corpus Benchmark
arXiv 2610.04636 · 2026-10-06LatentIndex：用于稀疏注意力的跨层共享与逐层选择
LatentIndex: Cross-Layer Sharing with Layer-Specific Selection for Sparse Attention
arXiv 2610.04635 · 2026-10-06奇异的 S^2 × S^2 候选者其实只是假的
Exotic S^2 x S^2 candidates that are just fake
arXiv 2610.04634 · 2026-10-06异质群体全基因组关联研究的统一无监督框架
A Unified Unsupervised Framework for Genome-Wide Association Studies in Heterogeneous Populations
arXiv 2610.04633 · 2026-10-06关于一类具有非局部速度的多维输运方程的注记
Remarks on a multi-dimensional transport equation with nonlocal velocity
arXiv 2610.04632 · 2026-10-06大型语言模型的数值线性代数
The Numerical Linear Algebra of Large Language Models
arXiv 2610.04631 · 2026-10-06基于损耗因子的多用户多中继系统能耗最小化
Minimizing Energy Consumption based on Waste Factor in Multi-user Multi-relay Systems
arXiv 2610.04630 · 2026-10-06非笛卡尔映射上的全球太阳可压缩对流模拟
Global solar compressible convection simulations on non-Cartesian mappings
arXiv 2610.04629 · 2026-10-06中心半直积上场论的协变Euler-Poincaré约化
Covariant Euler-Poincaré Reduction for Field Theories on Centered Semidirect Products
arXiv 2610.04628 · 2026-10-06图像退化下目标检测表示迁移的任务敏感几何
Task-Sensitive Geometry of Representation Transfer for Object Detection under Image Degradation
arXiv 2610.04627 · 2026-10-06流体天线安全信道驱动星座（FASCC）：一种物理层安全方法
Fluid Antenna Secure Channel-driven Constellation (FASCC): a Physical Layer Security approach
arXiv 2610.04625 · 2026-10-06稀疏观测PDE逆问题的后向一致扩散采样
Backward-Consistent Diffusion Sampling for Sparsely Observed PDE Inverse Problems
arXiv 2610.04624 · 2026-10-06子空间集中条件与Bogomolov--Gieseker不等式
Subspace concentration condition and Bogomolov--Gieseker inequality
arXiv 2610.04623 · 2026-10-06Campanato和Morrey空间的最优嵌入
Optimal Embeddings of Campanato and Morrey Spaces
arXiv 2610.04622 · 2026-10-06双侧局部随机化经典阴影的精确 Clifford 最优性
Exact Clifford Optimality for Two-Sided Locally Randomized Classical Shadows
arXiv 2610.04621 · 2026-10-06立场漂移：AI中介沟通如何扭曲我们的信息
Stance Drift: How AI-mediated Communication Distorts Our Message
arXiv 2610.04620 · 2026-10-06MUTACO：基于兼容性感知序贯设计的随机非线性系统中的模拟器高效响应匹配
MUTACO: Simulator-Efficient Response Matching in Stochastic Nonlinear Systems via Compatibility-Aware Sequential Design
arXiv 2610.04619 · 2026-10-06微波磁场的时空成像：基于磁振子相干劈裂
Spatiotemporal imaging of microwave magnetic fields via magnonic coherent splitting
arXiv 2610.04618 · 2026-10-06从图和拟阵构造经典码与量子码的新方法
New Methods for Constructing Classical and Quantum Codes from Graphs and Matroids
arXiv 2610.04617 · 2026-10-06PerturBot：通过扰动训练打破视觉-语言-动作模型中的捷径先验
PerturBot: Breaking Shortcut Priors in Vision-Language-Action Models with Perturbative Training
arXiv 2610.04616 · 2026-10-06PonyTail：剖析并改进服务尾部延迟
PonyTail: Profiling and Improving Service Tail Latency
arXiv 2610.04615 · 2026-10-06约束微分博弈中基于模型预测控制的近似反馈纳什均衡及其上界保证
Approximate Feedback Nash Equilibria in Constrained Differential Games via Model Predictive Control with Upper Bound Guarantees
arXiv 2610.04614 · 2026-10-06多轭表面码
Multi-Yoked Surface Codes
arXiv 2610.04613 · 2026-10-06基于可微有限元方法的软体气动机器人端到端建模框架
An End-to-End Framework for Modelling Pneumatic Soft Robots Based on Differentiable Finite Element Methods
arXiv 2610.04612 · 2026-10-06一阶时间需求的实际运行时执行
Practical Runtime Enforcement of First-Order Temporal Requirements
arXiv 2610.04611 · 2026-10-06非凸情形下的张量包络
Tensor Envelopes in the Nonconvex Setting
arXiv 2610.04610 · 2026-10-06利用层级控制器结构进行情境参数学习以实现人形机器人移动操作
Exploiting Hierarchical Controller Structure in Contextual Parameter Learning for Humanoid Loco-Manipulation
arXiv 2610.04609 · 2026-10-06一族二维Banach空间上的唯一算子空间结构
Unique operator space structures on a family of two dimensional Banach spaces
arXiv 2610.04608 · 2026-10-06ForeAct3D：面向VLA策略的策略接地未来世界建模
ForeAct3D: Policy-Grounded Future World Modeling for VLA Policies
arXiv 2610.04607 · 2026-10-06ConEx：通过概念感知归因实现人类可解释的显著性图
ConEx: Human-Interpretable Saliency Maps via Concept-Aware Attribution
arXiv 2610.04605 · 2026-10-06用大语言模型预测课程决策的后果
Anticipating the Consequences of Curriculum Decisions with Large Language Models
arXiv 2610.04604 · 2026-10-06
