2026-10论文中文摘要
pass@k 无法衡量的：训练后多样性与能力保持的评估
What pass@k Cannot Measure: Evaluating Diversity and Capability Retention after Post-Training
arXiv 2610.07405 · 2026-10-07超越一维的凸序：投影检验、反例与高斯混合
Convex Order Beyond Dimension One: Projection Tests, Counterexamples and Gaussian Mixtures
arXiv 2610.07404 · 2026-10-07深度防御大语言模型：评估记忆门控对抗激活诱导与记忆诱导的谄媚行为
Defense-in-Depth for LLMs: Evaluating Memory Gates Against Activation-Induced and Memory-Induced Sycophancy
arXiv 2610.07403 · 2026-10-07重新思考生成式推荐中的语义ID构建：SimHash与并行解码及语义对齐
Rethinking Semantic ID Construction for Generative Recommendation: SimHash with Parallel Decoding and Semantic Alignment
arXiv 2610.07402 · 2026-10-07$\varkappa$-Fréchet--Urysohn 与 Baire 子群在自由拓扑群中的性质
$\varkappa$-Fréchet--Urysohn and Baire subgroups of free topological groups
arXiv 2610.07401 · 2026-10-07Born-Oppenheimer有效场论：所有XYZ奇特态的统一框架
Born-Oppenheimer Effective Field Theory as a unified framework for all the XYZ exotics
arXiv 2610.07400 · 2026-10-07Fed-BRDECS：隐私保护与异构感知的联邦深度嵌入聚类
Fed-BRDECS: Privacy-Preserving and Heterogeneity-Aware Federated Deep Embedded Clustering
arXiv 2610.07399 · 2026-10-07自主、3D打印、喷水推进、开源机器人三体船用于环境巡检与监测
An Autonomous, 3D Printed, Waterjet-Powered, Open-Source Robotic Trimaran for Environmental Inspection and Monitoring
arXiv 2610.07398 · 2026-10-07引用是滞后的：从论文所相信的内容中解读认识论不稳定性，早于引用图追上数年
Citations Are Late: Reading epistemic instability from what papers believe, years before the citation graph catches up
arXiv 2610.07397 · 2026-10-07高程图无法看到的：面向人形机器人的语义感知运动与执行感知导航
What the Elevation Map Cannot See: Semantic-Aware Locomotion and Execution-Aware Navigation for Humanoid Robot
arXiv 2610.07396 · 2026-10-07追踪S型复合近地天体的主带起源：从空间任务目标到陨石来源
Tracing the main belt origins of S-complex near-Earth objects: From space mission targets to meteorite sources
arXiv 2610.07395 · 2026-10-07结合方案中的近因子分解
Near-factorizations in association schemes
arXiv 2610.07394 · 2026-10-0748自旋单重态流形的交换控制基准测试
Benchmarking exchange-only control of a 48-spin singlet manifold
arXiv 2610.07393 · 2026-10-07高维矩阵时间序列均值与协方差的多个变点检测
Multiple Change Point Detection in the Mean and Covariance of High-Dimensional Matrix Time Series
arXiv 2610.07392 · 2026-10-07关于使用聚合归一化流链进行复杂模拟模型似然逼近与推断的视角性注记
A perspective note on likelihood approximation and inference for complex simulation models using a chain of aggregated normalizing flows
arXiv 2610.07391 · 2026-10-07AeroBuoy：一种无人机可部署、3D打印的自主机器人浮标，用于偏远和危险河流系统的环境巡检
AeroBuoy: A Drone Deployable, 3D Printed, Autonomous Robotic Buoy for Environmental Inspection in Remote and Hazardous River Systems
arXiv 2610.07390 · 2026-10-07稀疏自编码器中的推理与学习作为自然梯度流
Inference and learning in sparse autoencoders as natural gradient flow
arXiv 2610.07389 · 2026-10-07关于拟阵约束下子模最大化的更强硬度结果
Stronger Hardness for Submodular Maximization Subject to a Matroid Constraint
arXiv 2610.07387 · 2026-10-07NetAgent：实用的多任务智能体网络流量分析
NetAgent: Multi-Task Agentic Network Traffic Analysis Made Practical
arXiv 2610.07386 · 2026-10-07视觉-语言模型微调过程中的语义能力获取与特化
Semantic Capability Acquisition and Specialization During Vision-Language Model Fine-Tuning
arXiv 2610.07385 · 2026-10-07WildMatch: 野生动物再识别的弱监督图像匹配器自适应
WildMatch: Weakly Supervised Image Matcher Adaptation for Wildlife Re-Identification
arXiv 2610.07384 · 2026-10-07HyperNSDE：用于联合静态-纵向临床数据生成的个性化神经随机微分方程
HyperNSDE: Personalized Neural SDEs for Joint Static-Longitudinal Clinical Data Generation
arXiv 2610.07383 · 2026-10-07迈向可信的物理AI以实现人机交互
Toward Trustworthy Physical AI for Human Interaction
arXiv 2610.07382 · 2026-10-07GeoWM：显式几何中的高效直接世界建模
GeoWM: Efficient Direct World Modeling in Explicit Geometry
arXiv 2610.07381 · 2026-10-07马尔可夫调制加性泛函的定量平均与Wentzell边界均匀化
Quantitative averaging of Markov-modulated additive functionals and Wentzell boundary homogenization
arXiv 2610.07380 · 2026-10-07随机非均匀超图的Property B
Property B for random non-uniform hypergraphs
arXiv 2610.07379 · 2026-10-07SimCortex v2：近零碰撞与自交的联合皮层表面重建
SimCortex v2: Joint Cortical Surface Reconstruction with Near-Zero Collisions and Self-Intersections
arXiv 2610.07378 · 2026-10-07开放云测试平台中面向科学计算的高速网络处理
High Speed Network Processing for Scientific Computing in the Open Cloud Testbed
arXiv 2610.07377 · 2026-10-07MemCo：面向泛化LLM智能体到未见环境的内存中心协作框架
MemCo: Memory-Centric Collaboration for Generalizing LLM Agents to Unseen Environments
arXiv 2610.07376 · 2026-10-07关于eutactic形式
On eutactic forms
arXiv 2610.07375 · 2026-10-07多组公平性与全知预测：分离与等价
Multigroup Fairness and Omniprediction: Separations and Equivalences
arXiv 2610.07374 · 2026-10-07粒子群优化的连续时间极限
A continuous-time limit for particle swarm optimization
arXiv 2610.07373 · 2026-10-07模型摩擦针织物的力学
Mechanics of a Model Frictional Knitted Fabric
arXiv 2610.07372 · 2026-10-07速度障碍与最近接近点度量中的表达性、等价性与不确定性
Expressiveness, Equivalence, and Uncertainty in Velocity Obstacles and Closest Point of Approach Metrics
arXiv 2610.07371 · 2026-10-07覆盖深度问题的最优码与随机码的性能
Optimal Codes for the Coverage Depth Problem and the Performance of Random Codes
arXiv 2610.07370 · 2026-10-07精确安全MPPI：基于非光滑控制障碍函数的安全感知采样
Exact-Safe MPPI: Safety-Aware Sampling with Nonsmooth Control Barrier Functions
arXiv 2610.07369 · 2026-10-07拜占庭容错的因果单播：恒定消息空间开销
Byzantine-Tolerant Causal Unicast with Constant Message Space Overhead
arXiv 2610.07368 · 2026-10-07奥卡姆剃刀与三族超轻费米子暗物质模型
Occam's Razor and a Three-Family Ultralight Fermionic Dark Matter Model
arXiv 2610.07367 · 2026-10-07身份条件化分数融合用于开放集行人重识别
Identity-Conditioned Score Fusion for Open-Set Person Re-Identification
arXiv 2610.07366 · 2026-10-07谁写的还不够：检测洞察力的贡献者
Who Wrote It Is Not Enough: Detecting Who Contributed the Insight
arXiv 2610.07365 · 2026-10-07SPECULATE：用于吸积盘风的合成光谱库、仿真器与拟合软件包——基础设施及其在CV中的应用
SPECULATE: A synthetic spectral library, emulator and fitting package for accretion disc winds -- Infrastructure and application to CVs
arXiv 2610.07364 · 2026-10-07边处理与节点结果的网络实验
Network Experiments with Edge Treatments and Node Outcomes
arXiv 2610.07363 · 2026-10-07硬资源约束下的大语言模型评估动态预算分配
Dynamic Budget Allocation for LLM Evaluation under Hard Resource Constraints
arXiv 2610.07362 · 2026-10-07中红外气泡的普适厚度
The universal thickness of mid-infrared bubbles
arXiv 2610.07361 · 2026-10-07通过 Blackwell 序研究多元高斯分布中的冗余与协同
Redundancy and synergy in multivariate Gaussians via the Blackwell order
arXiv 2610.07360 · 2026-10-07评估堆栈而非层：用于智能体动作的确定性与LLM门控是否独立失效？
Evaluate the Stack, Not the Layer: Do Deterministic and LLM Gates for Agent Actions Fail Independently?
arXiv 2610.07359 · 2026-10-07面向数据驱动的野火后泥石流预测的可解释基准测试
Towards Explainable Benchmarking for Data-driven Post-Wildfire Debris Flow Prediction
arXiv 2610.07358 · 2026-10-07ALMA 带通校准异常的分类器框架
A Classifier Framework for ALMA Bandpass Calibration Anomalies
arXiv 2610.07357 · 2026-10-07一个用于连贯多图SysML模型的有效数据集与基准
A Validated Dataset and Benchmark for Coherent Multi-Diagram SysML Models
arXiv 2610.07356 · 2026-10-07追踪并非持久性：视频世界模型对隐藏物体保留了什么
Tracking Is Not Permanence: What Video World Models Keep of a Hidden Object
arXiv 2610.07355 · 2026-10-07
