2026-10论文中文摘要
理解并缓解视觉语言模型中由Token剪枝引发的安全漏洞
Understanding and Mitigating Token-Pruning-Induced Vulnerabilities in VLMs
arXiv 2610.09703 · 2026-10-08生成式编辑下的潜在水印：检测存活的基准与分析
Latent Watermarks under Generative Editing: A Benchmark and Analysis of Detection Survival
arXiv 2610.09702 · 2026-10-08球谐域中保方向的空間音频信号增强
Direction-preserving Spatial Audio Signal Enhancement in The Spherical Harmonic Domain
arXiv 2610.09701 · 2026-10-08什么使得合成困难负样本在视觉-语言预训练中有效？
What Makes Synthetic Hard Negatives Work in Vision-Language Pretraining?
arXiv 2610.09700 · 2026-10-08模块化CMA-ES：现代进化策略的框架
The Modular CMA-ES: A Framework for Modern Evolution Strategies
arXiv 2610.09699 · 2026-10-08全$(2,q)$-正性中的虚假零色散连续谱：来自一个精确可解计数算子的线性层级深度下界
Spurious zero-dispersion continua in full $(2,q)$-positivity: A linear hierarchy-depth lower bound from an exactly solvable counting operator
arXiv 2610.09698 · 2026-10-08汽车硬件攻击：架构师的TARA指南
Automotive Hardware Attacks: An Architect's Guide to TARA
arXiv 2610.09697 · 2026-10-08RoboPace：面向动作块策略的接触感知时间最优重定时
RoboPace: Contact-Aware Time-Optimal Retiming for Action-Chunk Policies
arXiv 2610.09696 · 2026-10-08更好的视觉表征是否总能带来更好的端到端自动驾驶？
Do Better Visual Representations Always Lead to Better End-to-End Autonomous Driving?
arXiv 2610.09695 · 2026-10-08曲线的模空间的上同调欧拉特征
Motivic Euler characteristics of moduli spaces of curves
arXiv 2610.09694 · 2026-10-08两类错误的故事：探索自动化评判者在不确定性量化器评估中的隐性权衡
A Tale of Two Error Categories: Exploring Concealed Trade-Offs in the Errors of Automated Judges in Evaluation of Uncertainty Quantifiers
arXiv 2610.09693 · 2026-10-08面向个性化自动疼痛评估的少样本学习
Few-Shot Learning for Personalised Automated Pain Assessment
arXiv 2610.09692 · 2026-10-08限价订单簿下在线做市的最优遗憾
Optimal Regret for Online Market Making with Limit Order Book
arXiv 2610.09691 · 2026-10-08NAViLoss：一种用于物理一致性学习的水下导航感知双残差目标
NAViLoss: An Underwater Navigation-Aware Dual-Residual Objective for Physics-Consistent Learning
arXiv 2610.09690 · 2026-10-08二维无理环面上的Strichartz估计
Strichartz estimates on two-dimensional irrational tori
arXiv 2610.09689 · 2026-10-08次黎曼流形上最优传输的稳定性
Stability of optimal transport on sub-Riemannian manifolds
arXiv 2610.09688 · 2026-10-08轮廓算子：一维投影下低秩测度的可辨识性
The Silhouette Operator: Identifiability of Low-Rank Measures from One-Dimensional Projections
arXiv 2610.09687 · 2026-10-08精确球面三角形上的各向异性拉格朗日插值
Anisotropic Lagrange interpolation on exact spherical triangles
arXiv 2610.09686 · 2026-10-08DFT对固液界面的准确性如何？与随机相位近似在石墨烯-水界面的比较
How Good Is DFT for Solid-Liquid Interfaces? A Comparison With the Random-Phase Approximation for Water on Graphene
arXiv 2610.09685 · 2026-10-08从帕累托到偏好：通过摊销智能体策略发现实现个性化测试时扩展
From Pareto to Preference: Personalized Test-Time Scaling via Amortized Agentic Policy Discovery
arXiv 2610.09684 · 2026-10-08系统切换：快速决策模型何时应停下来思考？
System Switch: When Should a Fast Decision Model Stop and Think?
arXiv 2610.09683 · 2026-10-08基于传感器自身运动的自监督在空气声纳图像去卷积
Self-Supervised Deconvolution of In-Air Sonar Images Using Sensor Ego-Motion
arXiv 2610.09682 · 2026-10-08程序验证中基于智能体编排的成本高效定理证明
Cost-Efficient Theorem Proving via Agent Orchestration in Program Verification
arXiv 2610.09681 · 2026-10-08NNLO 初态修正对 $e^+\\,e^-\to\mu^+\\,\mu^-\\,\gamma$ 过程的影响
NNLO initial-state corrections to $e^+\,e^-\toμ^+\,μ^-\,γ$
arXiv 2610.09680 · 2026-10-08CERO：强化学习后训练中何时何地分配回放预算
CERO: Where and When to Allocate Rollouts for RL Post-Training
arXiv 2610.09679 · 2026-10-08KMS-对称量子马尔可夫半群的生成元的微分结构
The Differential Structure of Generators of KMS-Symmetric Quantum Markov Semigroups
arXiv 2610.09678 · 2026-10-08多源超声基准揭示当代自监督异常检测方法的局限性
A Multi-Source Ultrasound Benchmark Revealing the Limits of Contemporary Self-Supervised Anomaly Detection Methods
arXiv 2610.09677 · 2026-10-08黑洞冕的伽马射线：解析轻子与强子贡献
Gamma-rays from black hole coronae: disentangling the leptonic and hadronic contributions
arXiv 2610.09676 · 2026-10-08Gauss-Newton精度与不定Hessian：低成本集合中的一致共存
Gauss-Newton Accuracy and Indefinite Hessians: Uniform Coexistence in Low-Cost Sets
arXiv 2610.09675 · 2026-10-08基于物理接地合成预训练在有限实验数据下的实时托卡马克平衡重建
Real-time Tokamak Equilibrium Reconstruction Under Limited Experimental Data via Physics-Grounded Synthetic Pre-training
arXiv 2610.09674 · 2026-10-08通过动量选择性层析成像探索拓扑
Exploring topology via momentum-selective tomography
arXiv 2610.09673 · 2026-10-08有理点与 Hermitian-相对曲线上的拐点
Rational points and inflexions on Hermitian-relative curves
arXiv 2610.09672 · 2026-10-08InsClaimBench：跨决策链的保险理赔裁定基准测试
InsClaimBench: Benchmarking Insurance Claim Adjudication Across the Decision Chain
arXiv 2610.09671 · 2026-10-08准二值化自编码器：一种与架构无关的医学图像异常检测信息瓶颈
Quasi-Binarized Autoencoders: An Architecture-Independent Information Bottleneck for Medical Image Anomaly Detection
arXiv 2610.09670 · 2026-10-08一般最优停时的对偶刻画与算法
A dual characterization and algorithm of general optimal stopping
arXiv 2610.09669 · 2026-10-08剖面浮标与表层漂流浮标的统计海洋学
Statistical Oceanography of Profiling Floats and Surface Drifters
arXiv 2610.09668 · 2026-10-08共享与结构化输入削弱推理型AI智能体的集体随机选择
Shared and structured inputs undermine collective random choice by reasoning AI agents
arXiv 2610.09667 · 2026-10-08基于测地线的最优控制用于超导量子比特泄漏抑制
Geodesic-based optimal control for leakage suppression in superconducting qubits
arXiv 2610.09666 · 2026-10-08SAPD：步骤对齐的特权蒸馏
SAPD: Step-Aligned Privileged Distillation
arXiv 2610.09665 · 2026-10-08CrSb二维极限下的交替磁性消失与铁磁性涌现
Vanishing Altermagnetism and Emergent Ferromagnetism at the Two-Dimensional Limit of CrSb
arXiv 2610.09664 · 2026-10-08深度学习在大型地震干扰衰减中的实际应用：安哥拉Camie油田案例研究
Real-world application of deep learning in large-scale seismic interference attenuation: A case study in the Camie field of Angola
arXiv 2610.09663 · 2026-10-08封闭系统中渐近量子多体疤痕与开放系统中扩散性南部-戈德斯通模之间的对应关系
Correspondence between Asymptotic Quantum Many-Body Scars in Closed Systems and Diffusive Nambu-Goldstone Modes in Open Systems
arXiv 2610.09662 · 2026-10-08Alice：一个基于评分标准的多维度自动简答题评分的大规模德语基准
Alice: A Large-Scale German Benchmark for Rubric-Based Multi-Dimensional Automatic Short Answer Scoring
arXiv 2610.09661 · 2026-10-08Rubric Spans 是标签表示：用于短答案评分的联合 LLM 编码
Rubric Spans are Label Representations: Joint LLM Encoding for Short Answer Scoring
arXiv 2610.09660 · 2026-10-08面向分布式态矢量模拟的通信感知量子比特布局与自动节点数分配
Communication-Aware Qubit Placement and Automatic Node-Count Allocation for Distributed State-Vector Simulation
arXiv 2610.09659 · 2026-10-08无目标结局时生存模型比较的锐化部分识别
Sharp Partial Identification for Survival Model Comparison Without Target Outcomes
arXiv 2610.09658 · 2026-10-08快速且内存高效的混合XPU计算卸载训练框架
Fast and Memory Efficient Offload Training Framework with Hybrid XPU Computation
arXiv 2610.09657 · 2026-10-08统一处理关联金属中局域动力学相互作用的 Eliashberg 理论方法
Unified treatment of local dynamical interactions in correlated metals using Eliashberg theory
arXiv 2610.09656 · 2026-10-08一种基于约束递归训练的轻量级混合声学回声消除与抑制框架
A Lightweight Hybrid Framework for Acoustic Echo Cancellation and Suppression via Constrained Recursive Training
arXiv 2610.09655 · 2026-10-08DSTNet：用于因果多时域金融预测的动态谱轨迹网络
DSTNet: Dynamic Spectral Trajectory Network for Causal Multi-Horizon Financial Forecasting
arXiv 2610.09654 · 2026-10-08
