2026-09论文中文摘要
Blackwell 边界
Blackwell Boundaries
arXiv 2609.35062 · 2026-09-29恒定功率和低功率纠错冷却码的改进界
Improved bounds for constant-power and low-power error-correcting cooling codes
arXiv 2609.35061 · 2026-09-29超越梯度流：从分布快照中的可辨识性与恢复
Beyond Gradient Flow: Identifiability and Recovery from Distribution Snapshots
arXiv 2609.35060 · 2026-09-29面向可泛化三维异常检测的关系不一致性建模
Towards Generalizable 3D Anomaly Detection via Relational Inconsistency Modeling
arXiv 2609.35059 · 2026-09-29TIDE：通过信息蒸馏与探索实现智能体强化学习的师生过渡
TIDE: Teacher-Student Transition via Informative Distillation and Exploration for Agentic RL
arXiv 2609.35058 · 2026-09-29GRAVITY+时代的MATISSE
MATISSE in the era of GRAVITY+
arXiv 2609.35057 · 2026-09-29脑电解码的缩放定律：多少数据才够？
Scaling Laws for EEG Decoding: How Much Data Is Enough?
arXiv 2609.35056 · 2026-09-29因果Transformer在上下文长度上的普适性与泛化性
Universality and Generalization of Causal Transformers Across Context Lengths
arXiv 2609.35055 · 2026-09-29感知质量损失还是感知质量的损失？
Perceptual Quality Loss or Loss of Perceptual Quality?
arXiv 2609.35054 · 2026-09-29OPIS：视频世界模型中多对象记忆的输入锚定基准
OPIS: An Input-Grounded Benchmark for Multi-Object Memory in Video World Models
arXiv 2609.35052 · 2026-09-29亲吻数构造的大规模自主发现
Large-Scale Autonomous Discovery of Kissing Number Constructions
arXiv 2609.35051 · 2026-09-29高效处理几何奇异性的浸没边界法：应用于肋条
Handling geometric singularities efficiently with the immersed-boundary method: application to riblets
arXiv 2609.35050 · 2026-09-29万全之策：基于多管道融合的引力波瞬变探测
The best of all worlds: gravitational-wave transient detection with multiple pipelines
arXiv 2609.35049 · 2026-09-29通过拉格朗日移动构造不可分解的协边
Non-decomposable cobordisms via Lagrangian moves
arXiv 2609.35048 · 2026-09-29EMPIRIC：基于实验驱动的残差世界模型学习用于机器人规划
EMPIRIC: Experiment-Driven Learning of Residual World Models for Robot Planning
arXiv 2609.35047 · 2026-09-29LEGAU：学习语义高斯先验用于可扩展的类别级姿态估计
LEGAU: Learning Semantic Gaussian Priors for Scalable Category-level Pose Estimation
arXiv 2609.35046 · 2026-09-29杰斐逊实验室的介子光生：从双介子末态到介子-重子自旋密度矩阵
Meson photoproduction at Jefferson Lab: from two-meson final states to meson-baryon spin-density matrices
arXiv 2609.35045 · 2026-09-29BA-DPO：用于语言模型对齐的偏差调整直接偏好优化
BA-DPO: Bias-Adjusted Direct Preference Optimization for Language Model Alignment
arXiv 2609.35044 · 2026-09-29混合先验决策风险用于开放集识别
Mixed-Prior Decision Risk for Open-Set Recognition
arXiv 2609.35043 · 2026-09-29基于图学习的多时间尺度火星大气预测
Graph-Based Learning for Multi-Horizon Martian Atmospheric Forecasting
arXiv 2609.35042 · 2026-09-29GAZEleak：通过外部观测针对眼动追踪XR设备的密码推断攻击
GAZEleak: Passcode Inference Against Eye-tracking XR Devices Through External Observation
arXiv 2609.35040 · 2026-09-29不确定时不要切割：用于机器人采摘中VLA策略的可拒绝且校准的决策头
Do Not Cut When Uncertain: Rejectable and Calibrated Decision Heads for VLA Policies in Robotic Harvesting
arXiv 2609.35039 · 2026-09-29GUIDE-FBO：通过不确定性干预与分布交换实现联邦贝叶斯优化的引导
GUIDE-FBO: Guidance via Uncertainty Intervention and Distributional Exchange for Federated Bayesian Optimization
arXiv 2609.35038 · 2026-09-29多元正态性BHEP检验的四矩近似与快速$p$值计算
Four-Moment Approximations and Fast $p$-Values for BHEP Tests of Multivariate Normality
arXiv 2609.35037 · 2026-09-29人格跟随并非选择性控制：大语言模型用户模拟中的中性差距
Persona Following Is Not Selective Control: The Neutrality Gap in LLM User Simulation
arXiv 2609.35036 · 2026-09-29基于检验-相关性分解的排名相关检验下推荐排序离线策略评估
Recommendation Ranking Off-Policy Evaluation under Ranking-Dependent Examination via Examination-Relevance Decomposition
arXiv 2609.35034 · 2026-09-29关于有限喷管中带热传导的一维稳态Euler系统在热传导系数趋于零时的奇异层
On the singular layer for the 1-D steady Euler system with heat conductivity in a finite nozzle as the heat conduction coefficient vanishes
arXiv 2609.35033 · 2026-09-29JRDB-AVR：真实世界环境中具身智能体的主动视觉推理基准
JRDB-AVR: An Active Visual Reasoning Benchmark for Embodied Agents in Real-World Environments
arXiv 2609.35032 · 2026-09-29PEAR：面向工业搜索系统的渐进式基于证据的自动研究
PEAR: Progressive Evidence-Based AutoResearch for Industrial Search Systems
arXiv 2609.35031 · 2026-09-29大质量伪相对论非线性薛定谔方程正径向基态的唯一性
Uniqueness of positive radial ground states for the massive pseudo-relativistic nonlinear Schrödinger equation
arXiv 2609.35030 · 2026-09-29浅层线性网络在增长训练时域下的快速学习率迁移
Fast Learning Rate Transfer in Shallow Linear Networks at Growing Training Horizons
arXiv 2609.35029 · 2026-09-29VEX-Bench：评估LLM生成虚假信息的验证复杂度基准
VEX-Bench: Benchmarking Verification Complexity of LLM-Generated Misinformation
arXiv 2609.35028 · 2026-09-29计算欧几里得格覆盖半径的复杂性
The complexity of computing the covering radius of a Euclidean lattice
arXiv 2609.35027 · 2026-09-29WebPageBench：面向Web智能体的事件级验证与受控UI变体生成
WebPageBench: Event-Level Verification and Controlled UI-Variant Generation for Web Agents
arXiv 2609.35026 · 2026-09-29AutoDataBench：智能体能否编写驱动自我改进循环的数据？
AutoDataBench: Can Agents Write the Data That Feeds the Self-Improvement Loop?
arXiv 2609.35025 · 2026-09-29关于SYK模型的本征向量
On eigenvectors of the SYK model
arXiv 2609.35024 · 2026-09-29Proxy2World：从轻量级代理学习生成世界而无需看到它们
Proxy2World: Learning to Generate Worlds From Lightweight Proxies without Seeing Them
arXiv 2609.35023 · 2026-09-29基于频谱特征的超分辨率模型对抗攻击检测
Detection of Adversarial Attacks on Super-Resolvers Using Spectral Features
arXiv 2609.35022 · 2026-09-29通过最优传输引导的掩码解决时空预测中的空间不可区分性
Addressing Spatial Indistinguishability in Spatiotemporal Prediction via Optimal Transport-Guided Masking
arXiv 2609.35021 · 2026-09-29验证线性表示假说：视觉SAE的可解释性如何？
Verifying the Linear Representation Hypothesis: How Interpretable Are Vision SAEs?
arXiv 2609.35020 · 2026-09-29复高斯 Hafnian 的反集中性
Anticoncentration of Complex Gaussian Hafnians
arXiv 2609.35019 · 2026-09-29使用进化可塑性人工神经网络的智能体利用社会信息的环境要求
Environmental requirements for the use of social information by artificial life agents using evolved plastic artificial neural networks
arXiv 2609.35018 · 2026-09-29TermJudge：一种文档级指标，用于评判而非计数机器翻译评估中的术语
TermJudge: A Document-Level Metric Judging, Not Counting, Terminology in Machine Translation Evaluation
arXiv 2609.35017 · 2026-09-29利用Grover算法的实验性量子计算增强传感
Experimental quantum-computing-enhanced sensing using Grover's algorithm
arXiv 2609.35016 · 2026-09-29遍历性、遍历性破缺及非平衡态下概率密度的时间演化：多散斑动态光散射的新视角
Ergodicity, ergodicity breaking, and temporal evolution of probability density in non-equilibrium: new perspectives by multispeckle DLS
arXiv 2609.35014 · 2026-09-29摊销反馈规划：将基于模型的展开转化为可执行策略
Amortized Feedback Planning: Turning Model-Based Rollouts into Executable Policies
arXiv 2609.35012 · 2026-09-29欧盟价格稳定性：基于随机矩阵理论的系统性方法
Price Stability in the European Union: A Systemic Approach Using Random Matrix Theory
arXiv 2609.35011 · 2026-09-29面向可旋转天线赋能低空经济网络的用户关联与发射波束成形
User Association and Transmit Beamforming for Rotatable Antenna-Enabled LAE Networks
arXiv 2609.35010 · 2026-09-29加权奇异向量的豪斯多夫维数
Hausdorff Dimension of Weighted Singular Vectors
arXiv 2609.35009 · 2026-09-29手性分子在高$J$转动能态下的对映体特异性泵浦
Enantiomer-specific pumping of chiral molecules in high-$J$ rotational states
arXiv 2609.35008 · 2026-09-29
