2026-10论文中文摘要
WorldSolver：LLM智能体能否通过求解器生成来模拟物理动力学？
WorldSolver: Can LLM Agents Simulate the Physical Dynamics via Solver Generation?
arXiv 2610.08720 · 2026-10-07保留最佳N个（Holdout Best-of-N）：无偏评估及其代价
Holdout Best-of-N: Unbiased Evaluation and Its Cost
arXiv 2610.08719 · 2026-10-07当遗忘并非灾难性：关于虚假遗忘的机制
When Forgetting is not Catastrophic: On the Mechanics of Spurious Forgetting
arXiv 2610.08718 · 2026-10-07路径与流的协同演化：通过路径-流对齐
Co-Evolving Paths and Flows via Path-Flow Alignment
arXiv 2610.08717 · 2026-10-07解耦生成式检索中的范式、标识符与解码
Disentangling Paradigm, Identifier, and Decoding in Generative Retrieval
arXiv 2610.08716 · 2026-10-07面向跨空间时间序列的预测驱动推断
Prediction-powered inference for time series across space
arXiv 2610.08715 · 2026-10-07多变量多项式变换的量子算法：从高效合成到量子信道变换
Quantum Algorithms for Multivariable Polynomial Transformations: From Efficient Synthesis to Quantum Channel Transformations
arXiv 2610.08714 · 2026-10-07SpaTime：面向时空推理的流式视觉语言模型
SpaTime: Streaming Vision-Language Models for Spatio-temporal Reasoning
arXiv 2610.08713 · 2026-10-07大型网络中Moran's $I$和同配性的渐近零分布
Asymptotic Null Distributions of Moran's $I$ and Assortativity in Large Networks
arXiv 2610.08711 · 2026-10-07纠缠突然死亡：Eberly的工作与局域动力学的极限
Entanglement Sudden Death: Eberly's Work and the Limits of Local Dynamics
arXiv 2610.08710 · 2026-10-07宇宙时间尺度上大质量黑洞对宿主中的星系合并驱动特征
Galaxy merger-driven signatures in massive black hole pair hosts across cosmic time
arXiv 2610.08709 · 2026-10-07无拥挤超图中的随机独立集
Random independent sets in uncrowded hypergraphs
arXiv 2610.08707 · 2026-10-07非线性杆链模型：零能链与连续统一致性
A nonlinear rod linkage model: zero-energy chains and continuum consistency
arXiv 2610.08705 · 2026-10-07基于扩散模型的图像风格化中的局部内容-风格控制
Local Content-Style Control for Diffusion-based Image Stylization
arXiv 2610.08704 · 2026-10-07一致性并非有效性：K-12数学辅导对话中诊断学生失败模式的跨模型LLM共识
Agreement Is Not Validity: Cross-Model LLM Consensus in Diagnosing Student Failure Modes in K-12 Math Tutoring Dialogue
arXiv 2610.08703 · 2026-10-07基于场景的电池储能系统数据驱动预测控制鲁棒性研究：一项实验研究
Robust Scenario-Based Data-Enabled Predictive Control of a Battery Energy Storage System: An Experimental Study
arXiv 2610.08702 · 2026-10-07大规模增长会发生：随机矩阵上部分主元高斯消元法
Large Growth Happens: Gaussian Elimination with Partial Pivoting on Random Matrices
arXiv 2610.08701 · 2026-10-07reVISit-XR：将扩展现实引入可嵌入、可追踪、可回放的可视化研究
reVISit-XR: Bringing Extended Reality into Embeddable, Trackable, and Replayable Visualization Studies
arXiv 2610.08700 · 2026-10-07nanoMuse：面向你拥有的每一台设备的开源个人智能体
nanoMuse: An Open-Source Personal Agent for Every Device You Own
arXiv 2610.08699 · 2026-10-07PrimitiveCAD：基于LLM的点到CAD重建，采用基元感知标记化与操作对齐
PrimitiveCAD: An LLM-Based Point-to-CAD Reconstruction with Primitive-Aware Tokenization and Operation Alignment
arXiv 2610.08698 · 2026-10-07多输入离散时间平坦系统的精确线性化
On the Exact Linearization of Multi-Input Discrete-Time Flat Systems
arXiv 2610.08697 · 2026-10-07里德伯原子接收机用于集成感知与通信的基础CRB-速率权衡
Fundamental CRB-Rate Tradeoff of Rydberg Atomic Receivers for Integrated Sensing and Communication
arXiv 2610.08696 · 2026-10-07NMPP：杂乱环境中敏捷无人机飞行的非线性模型预测规划
NMPP: Nonlinear Model Predictive Planning for Agile UAV Flight in Cluttered Environments
arXiv 2610.08695 · 2026-10-07GeneICL：面向批量转录组学的表格基础模型
GeneICL: A Tabular Foundation Model for Bulk Transcriptomics
arXiv 2610.08694 · 2026-10-07神经算子预测的时间相关约化子空间用于非线性偏微分方程的基于投影的模拟
Neural-Operator-Predicted Time-Dependent Reduced Subspaces for Projection-Based Simulation of Nonlinear PDEs
arXiv 2610.08693 · 2026-10-07ScienceClaw：面向自然科学与社会科学中AI科学智能体的持续自进化基准测试
ScienceClaw: Benchmarking Continual Self-Evolution of AI-for-Science Agents Across the Natural and Social Sciences
arXiv 2610.08691 · 2026-10-07奇阶线性群的交换商与最大轨道大小
Abelian quotients and the largest orbit sizes of linear groups of odd order
arXiv 2610.08690 · 2026-10-07高斯过程因果模型中离散结果的反事实概率推断
Probabilistic Counterfactual Inference for Discrete Outcomes in Gaussian-Process Causal Models
arXiv 2610.08689 · 2026-10-07面向边缘AI的RISC-V Transformer推理加速框架
A Framework for Accelerating Transformer Inference on RISC-V for Edge AI
arXiv 2610.08688 · 2026-10-07单轴铁电体中拓扑极化结构的相场洞察
Phase-field insight into the nature of topological polarization structures in uniaxial ferroelectrics
arXiv 2610.08687 · 2026-10-07驯服可缩 $n$-流形的拓扑，$n=6,7$，及 BiRicci 曲率
Topology of tame contractible $n$-manifolds, $n=6,7$, and BiRicci Curvature
arXiv 2610.08686 · 2026-10-07刻画随机惯性：噪声何时会延长逃逸时间？
Characterizing stochastic inertia: when does noise increase escape times?
arXiv 2610.08685 · 2026-10-07RenderBench：基于重建数字孪生的渲染到真实视频迁移基准
RenderBench: Benchmarking Render-to-Real Video Transfer with Reconstructed Digital Twins
arXiv 2610.08684 · 2026-10-07耦合但滞后：全双工语音模型在无脚本对话中的话轮转换
Coupled but Late: Turn-Taking Between Full-Duplex Speech Models in Unscripted Dialogue
arXiv 2610.08683 · 2026-10-07超临界 Bernoulli 逾渗在 $\mathbb{Z}^d$（$d\geq 3$）上的非微扰有限连通性
Non-perturbative finite connectivities in supercritical Bernoulli percolation on $\mathbb{Z}^d$ with $d\geq 3$
arXiv 2610.08682 · 2026-10-07Juicy 交互式可视化：评估过度反馈设计如何塑造可视化参与度
Juicy Interactive Visualization: Evaluating How Excessive Feedback Design Shapes Visualization Engagement
arXiv 2610.08681 · 2026-10-07小语言模型在抽象推理任务上的系统性研究
A Systematic Study of Small Language Models on Abstract Reasoning Tasks
arXiv 2610.08680 · 2026-10-07二维Euler方程在单连通有界域中斑块结构持久性的几何效应
Geometric effects on the persistence of patch structures for 2D Euler equations in simply-connected bounded domains
arXiv 2610.08679 · 2026-10-07大型语言模型的安全投机解码
Secure Speculative Decoding for Large Language Models
arXiv 2610.08678 · 2026-10-07方差最优的带联合效应建模的离线策略评估
Variance-Optimal Off-Policy Evaluation with Conjunct Effect Modeling
arXiv 2610.08677 · 2026-10-07同数引用互换：压力测试 Jev 作为金融证据裁判
Same-Number Citation Swaps: Stress-Testing Jev as a Financial Evidence Judge
arXiv 2610.08675 · 2026-10-07EC-RAG：面向长视频理解的事件链检索增强生成
EC-RAG: Event Chain Retrieval-Augmented Generation for Long Video Understanding
arXiv 2610.08674 · 2026-10-07QCD_2：'t Hooft方程与相对论薛定谔方程
QCD_2: 't Hooft equation vs. relativistic Schroedinger equation
arXiv 2610.08673 · 2026-10-07PDB：基于点的形变混合用于面部动画重定向
PDB: Point-Based Deformation Blending for Facial Animation Retargeting
arXiv 2610.08672 · 2026-10-07增强几率：利用动量场进行宇称破缺的低阶探针
Boosted Odds: Lower-Point Probes of Parity Violation with Momentum Fields
arXiv 2610.08671 · 2026-10-07压力下的原则性：后训练决定大语言模型是否依据自身道德判断行事
Principled Under Pressure: Post-Training Decides Whether LLMs Act on Their Own Moral Judgment
arXiv 2610.08670 · 2026-10-07MemFLoRA：面向边缘CNN适配的内存下限LoRA
MemFLoRA: Memory-Floor LoRA for CNN Adaptation at the Edge
arXiv 2610.08669 · 2026-10-07语义行为水印：面向LLM智能体的释义鲁棒且防伪造溯源
Semantic Behavioral Watermarking: Paraphrase-Robust and Forgery-Resistant Provenance for LLM Agents
arXiv 2610.08668 · 2026-10-07基于Lindblad模拟的基态制备容错资源估计
Fault-tolerant resource estimation for ground-state preparation via Lindblad simulation
arXiv 2610.08667 · 2026-10-07南希·格蕾丝·罗曼太空望远镜宽视场仪器：热真空测试期间亮源光栅饱和响应
Nancy Grace Roman Space Telescope Wide Field Instrument: Grism Saturation Response to Bright Sources during Thermal-Vacuum Testing
arXiv 2610.08666 · 2026-10-07
