2026-10论文中文摘要
Vosti：指定、实现和验证确定性LLM推理
Vosti: Specifying, Implementing, and Verifying Deterministic LLM Inference
arXiv 2609.38981 · 2026-10-01SCORE-LM：基于语言模型的状态空间雷达表示用于故障诊断
SCORE-LM: State-Space Radar Representations with Language Models for Fault Diagnosis
arXiv 2609.38980 · 2026-10-01PARK：视频扩散Transformer中稀疏注意力的精确块检索
PARK: Accurate Block Retrieval for Sparse Attention in Video Diffusion Transformers
arXiv 2609.38978 · 2026-10-01尺度分裂神经算子用于内存高效和数据高效的三维湍流预测
Scale-Split Neural Operator for Memory- and Data-Efficient 3D Turbulence Prediction
arXiv 2609.38977 · 2026-10-01超越单次运行的公平性：语音大语言模型适配中的训练种子变异性
Fairness Beyond a Single Run: Training-Seed Variability in Speech LLM Adaptation
arXiv 2609.38976 · 2026-10-01非偶然性逻辑的证明论
Proof Theory for Non-Contingency Logic
arXiv 2609.38975 · 2026-10-01RealWorldShop：真实世界电子商务中对话式购物智能体的基准测试与改进
RealWorldShop: Benchmarking and Improving Conversational Shopping Agents in Real-World E-commerce
arXiv 2609.38974 · 2026-10-01多体谱边缘的本征态热化
Eigenstate thermalization at the edge of the many-body spectrum
arXiv 2609.38973 · 2026-10-01弯曲的拒绝：测量与预测具身视觉语言规划器中的任务可塑性
Refusals That Bend: Measuring and Predicting Task Malleability in Embodied VLM Planners
arXiv 2609.38971 · 2026-10-01小加法群满足 $A+ξA=\mathbb R$
Small additive groups with $A+ξA=\mathbb R$
arXiv 2609.38970 · 2026-10-01来自课堂实时投票的学生学习洞察：不止于对错
Insights on Student Learning from Live Classroom Polls: More Than Right or Wrong
arXiv 2609.38969 · 2026-10-01超越空间域监督：用于多模态图像融合的关系约束空间
Beyond Spatial-Domain Supervision: A Relation Constrained Space for Multi-Modal Image Fusion
arXiv 2609.38968 · 2026-10-01学生实际提出的问题：混合支持系统中的需求结构与自动化潜力
What Students Actually Ask: Demand Structure and Automation Potential in a Hybrid Support System
arXiv 2609.38967 · 2026-10-01Video2SwimFish：一种从真实鱼类视频重建可控鱼类模型与生物运动学的自动化流水线
Video2SwimFish: An Automated Pipeline for Reconstructing Controllable Fish Models and Biological Locomotion from Real Fish Videos
arXiv 2609.38966 · 2026-10-01一种用于具有高阶相互作用的玻色-爱因斯坦凝聚体基态的正守恒傅里叶优化方法
A positive-conservative Fourier optimization method for ground states of Bose--Einstein condensates with higher-order interactions
arXiv 2609.38965 · 2026-10-01顺序重要：多智能体顺序辩论中的首位发言者偏见及其通过人格特质的缓解
When Order Matters: First-Speaker Bias and Mitigation through Personality in Sequential Multi-Agent Debate
arXiv 2609.38964 · 2026-10-01Sartre事件生成器中小x eA碰撞的包容性衍射
Inclusive diffraction in eA at small x with the Sartre event generator
arXiv 2609.38963 · 2026-10-01缓解推理任务中的幻觉：基于无需训练的、不确定性引导的转向方法
Alleviating Hallucination in Reasoning Tasks with Training-Free Uncertainty-Guided Steering
arXiv 2609.38962 · 2026-10-01部分可积系统的绝热不变量作用量
Adiabatic Invariant Actions for Partially Integrable Systems
arXiv 2609.38961 · 2026-10-01基于分区最优传输的快速可扩展多智能体分布匹配
Fast and Scalable Multi-Agent Distribution Matching via Partitioned Optimal Transport
arXiv 2609.38960 · 2026-10-01大型课堂投票语料库中测量质量与答案偏见的审计
An Audit of Measurement Quality and Answer Bias in a Large Classroom-Poll Corpus
arXiv 2609.38959 · 2026-10-01定向检索、紧凑表示：CoT推理如何提升长上下文计数
Targeted Retrieval, Compact Representations: How CoT Reasoning Improves Long-Context Counting
arXiv 2609.38958 · 2026-10-01可交换局域哈密顿问题：对BQP困难性的相对化证据
The Commuting Local Hamiltonian Problem: Relativized Evidence Against BQP-Hardness
arXiv 2609.38957 · 2026-10-01路由探针可以在没有新信息的情况下改进：模型输出之外不确定性的精确零假设审计
Routing Probes Can Improve Without New Information: An Exact-Null Audit of Uncertainty Beyond Model Outputs
arXiv 2609.38956 · 2026-10-01无循环逆强化学习：通过Q分数匹配的顺序价值恢复
Loop-Free Inverse Reinforcement Learning via Sequential Value Recovery with Q-Score Matching
arXiv 2609.38955 · 2026-10-01APTInvestBench：评估不同遥测条件下的自主APT调查
APTInvestBench: Evaluating Autonomous APT Investigation under Varying Telemetry
arXiv 2609.38954 · 2026-10-01三角形上的尖锐诺伊曼特征值均值：精确证书证明与稳定性
Sharp Neumann eigenvalue means on triangles:an exact certificate proof and stability
arXiv 2609.38953 · 2026-10-01抛物均匀化中的奇异集与加倍性质
Singular Sets and Doubling Properties in Parabolic Homogenization
arXiv 2609.38952 · 2026-10-01测量大型第一年编程课程中学生自我评估与口试展示掌握程度的一致性
Measuring Student Self-Assessment against Viva-Demonstrated Mastery in a Large First-Year Programming Course
arXiv 2609.38951 · 2026-10-01区域尺度非线性结构响应模拟：基于建筑特定Bouc-Wen模型
Regional-scale nonlinear structural response simulation using building-specific Bouc-Wen models
arXiv 2609.38950 · 2026-10-01文本-视频检索：基于多维显著性评估与粒度感知查询分解
Text-Video Retrieval via Multi-Dimensional Saliency Assessment and Granularity-Aware Query Decomposition
arXiv 2609.38949 · 2026-10-01DrivingBench：视觉语言模型能否驾驶丰田卡罗拉？
DrivingBench: Can Vision-Language Models Drive a Toyota Corolla?
arXiv 2609.38948 · 2026-10-01打破过滤气泡的新闻：生成式AI搜索多样化集体注意力并提高共享信息消费
Breaking News Out of the Filter Bubble: Generative AI Search Diversifies Collective Attention and Raises Shared Information Consumption
arXiv 2609.38946 · 2026-10-01以最优速率认证随机性
Certified Randomness with Optimal Rate
arXiv 2609.38945 · 2026-10-01主动光机械腔中的频率梳产生
Frequency comb generation in active optomechanical cavities
arXiv 2609.38944 · 2026-10-01量子回波马尔可夫过程用于组合优化
Quantum-echo Markov process for combinatorial optimization
arXiv 2609.38943 · 2026-10-01二维CeN的第一性原理研究：结构稳定性、电子结构与磁性之间的相互作用
First-Principles Study of Two-Dimensional CeN: Interplay between Structural Stability, Electronic Structure, and Magnetism
arXiv 2609.38942 · 2026-10-01超重核合成中俘获截面的普适经验耦合道计算
Universal empirical coupled channel calculations for capture cross sections in synthesis of superheavy nuclei
arXiv 2609.38940 · 2026-10-01缺陷感知的并行原子重载协议用于中性原子量子计算机
Defect-Aware Parallel Atom Reloading Protocol for Neutral-Atom Quantum Computers
arXiv 2609.38939 · 2026-10-01鲁棒风险敏感强化学习：从被污染的人类反馈中学习
Robust Risk-Sensitive Reinforcement Learning from Corrupted Human Feedback
arXiv 2609.38938 · 2026-10-01湍流马赫数对恒星形成和初始质量函数的影响
The impact of the turbulent Mach number on star formation and the initial mass function
arXiv 2609.38937 · 2026-10-01信号路由温度缩放：低容量风险条件校准用于小验证预算
Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets
arXiv 2609.38936 · 2026-10-01PrivCert：在差分隐私下认证语句支持度
PrivCert: Certifying Statement Support under Differential Privacy
arXiv 2609.38934 · 2026-10-01EFormer：用于连续sEMG手部姿态跟踪的时间对齐局部校正
EFormer: Temporally Aligned Local Correction for Continuous sEMG-Based Hand Pose Tracking
arXiv 2609.38932 · 2026-10-01自适应自洽性：从黑盒采样到分布值反馈
Adaptive Self-Consistency: From Black-Box Sampling to Distribution-Valued Feedback
arXiv 2609.38931 · 2026-10-01图像分割中条件独立假设的松弛
On the Relaxation of Conditional Independence Assumption for Image Segmentation
arXiv 2609.38930 · 2026-10-01学习遗忘什么：LLM表示空间的分布遗忘
Learning What to Forget: Distributional Unlearning for LLM Representation Spaces
arXiv 2609.38929 · 2026-10-01局部最小值逃逸器：未知环境中鲁棒导航的程序化子目标生成
Local-Minimum Escaper: Programmatic Subgoal Generation for Robust Navigation in Unknown Environments
arXiv 2609.38928 · 2026-10-01世界即图：通过潜在空间图进行关系世界建模
World-as-Graph: Relational World Modeling Through Latent Space Graphs
arXiv 2609.38927 · 2026-10-01PrecipJEPA：基于运动源渲染的JEPA正则化未来状态预测用于降水临近预报
PrecipJEPA: JEPA-Regularized Future-State Prediction with Motion-Source Rendering for Precipitation Nowcasting
arXiv 2609.38926 · 2026-10-01
