2026-09论文中文摘要
OSFoundry：利用规范引导的智能体构建与演进操作系统
OSFoundry: Building and Evolving Operating Systems with Specification-Guided Agents
arXiv 2609.25018 · 2026-09-23深度伪造与合成媒体：生成、检测与治理
Deepfakes and Synthetic Media: Generation, Detection, and Governance
arXiv 2609.25017 · 2026-09-23Nielsen-Soelberg 群环的整环问题：交换情形，附认证球检验
The domain question for the Nielsen-Soelberg group rings: the commutative case, with certified ball checks
arXiv 2609.25016 · 2026-09-23整群环的扭曲酉元紧致性定理，以及通向Promislow群B(4)窗口theta-酉情形A的认证路径
A compactness theorem for twisted-unitary elements of integral group rings, with a certified route to the theta-unitary Case A at window B(4) of the Promislow group
arXiv 2609.25015 · 2026-09-23并非所有4位量化器都相同：微调小型语言模型中PII泄露的部署时缓解
Not All 4-bit Quantizers Are Equal: Deployment-Time Mitigation of PII Leakage in Fine-Tuned Small Language Models
arXiv 2609.25014 · 2026-09-23现有预条件器能否提升生物医学表格基础学习？TabPFN优化的实证研究
Do Existing Preconditioners Improve Biomedical Tabular Foundation Learning? An Empirical Study on TabPFN Optimization
arXiv 2609.25013 · 2026-09-23一种计算梵语文献中语义变化的方法
A Computational Approach to Measuring Semantic Change in Sanskrit Literature
arXiv 2609.25012 · 2026-09-23Erdős-Mollin-Walsh 猜想在 $10^{14}$ 以下的核验证
A Kernel-Certified Verification of the Erdős-Mollin-Walsh Conjecture below $10^{14}$
arXiv 2609.25011 · 2026-09-23合成人物能否预测真实受众反应？一项模拟到真实的研究：无人物基线优于基于人物的文案模拟
Do Synthetic Personas Predict Real Audience Response? A Sim-to-Real Study Where a No-Persona Baseline Beats Persona-Based Copy Simulation
arXiv 2609.25010 · 2026-09-23相同数量，不同答案：语言模型中的数值表示不变性
Same Quantity, Different Answer: Numerical Representation Invariance in Language Models
arXiv 2609.25009 · 2026-09-23用 Rust 端到端训练语言模型：一份经验报告
Training a Language Model End-to-End in Rust: An Experience Report
arXiv 2609.25008 · 2026-09-23超越短片段：利用向量档案扩展说话人嵌入
Beyond Short Segments : Expanding Speaker Embeddings with Vector Archives
arXiv 2609.25007 · 2026-09-2399%准确率衡量的是什么？对广泛使用的假新闻语料库中捷径学习的可复现审计
What Does 99% Accuracy Measure? A Reproducible Audit of Shortcut Learning in a Widely Used Fake News Corpus
arXiv 2609.25006 · 2026-09-23高温预处理对COTS LG Chem HG2电池低温循环过程中锂沉积的影响
Effects of High-Temperature Preconditioning on Li Plating During Low-Temperature Cycling of the COTS LG Chem HG2 Cell
arXiv 2609.25005 · 2026-09-23DolphinBench：绘制智能体记忆的帕累托前沿
DolphinBench: Mapping the Pareto Frontier of Agent Memory
arXiv 2609.24971 · 2026-09-23逼近1.556以下的奖赏收集旅行商问题
Approximating Prize-Collecting TSP below 1.556
arXiv 2609.24944 · 2026-09-23解剖分解的胸部计算机断层扫描（CT）投影作为胸部X光片中骨骼抑制的可扩展监督
Anatomy-Decomposed Chest Computed Tomography (CT) Projections as Scalable Supervision for Bone Suppression in Chest Radiographs
arXiv 2609.24937 · 2026-09-23玻色子星工厂：过去、现在与未来
Boson Star Factory: Past, Present, and Future
arXiv 2609.24913 · 2026-09-23Pinocchio：黑盒语言模型的快速不确定性估计
Pinocchio: Fast Uncertainty Estimates for Black-Box Language Models
arXiv 2609.24881 · 2026-09-23量子干涉驱动的轨道密度波与三层镍酸盐中的高温超导电性
Quantum-interference-driven orbital density wave and high-temperature superconductivity in trilayer nickelates
arXiv 2609.24857 · 2026-09-23用于医学诊断的移动成像解决方案：趋势与应用
Mobile Imaging Solutions for Medical Diagnosis: Trends and Applications
arXiv 2609.24814 · 2026-09-23色玻璃凝聚与共线框架在大横动量半单举深度非弹性散射中的比较研究
Comparative Study of Color Glass Condensate and Collinear Frameworks for Large-Transverse-Momentum Semi-Inclusive Deep Inelastic Scattering
arXiv 2609.24783 · 2026-09-23来自李群的临界维数中的自由浸入
Free Immersions in Critical Dimension from Lie Groups
arXiv 2609.24777 · 2026-09-23形式化 $f(R)$ 与 $f(Q)$ 引力的统一动力学系统框架
Formalizing a unified dynamical system framework for $f(R)$ and $f(Q)$ gravity
arXiv 2609.24733 · 2026-09-23固定点附近一个共振双全纯映射的扇形规范化
Sectorial normalisation of one resonant biholomorphisms near a fixed point
arXiv 2609.24695 · 2026-09-23像世界模型一样思考，像VLA一样行动：将世界模型表征蒸馏进紧凑型机器人策略
Think Like a World Model, Act Like a VLA: Distilling World-Model Representations into Compact Robot Policies
arXiv 2609.24682 · 2026-09-23Touch2Robot：人类演示循环中的机器人触觉
Touch2Robot: Robot Touch in the Human Demonstration Loop
arXiv 2609.24660 · 2026-09-23由受控粗糙路径驱动的粗糙微分方程生成的随机动力系统
Random dynamical systems generated by rough differential equations driven by controlled rough paths
arXiv 2609.24637 · 2026-09-23IGRINS 高转动振动H2观测揭示的光离解区物理条件
Physical conditions in PDRs revealed by IGRINS ro-vibrational H2 observations
arXiv 2609.24630 · 2026-09-23沿平坦态射下降与根式化复合
Descent along flat composed with radicalization
arXiv 2609.24577 · 2026-09-23超越子模性的泊松交换：离线与在线拟阵子集选择的有效近似算法
Poisson Exchange Beyond Submodularity: Effective Approximation Algorithms for Offline and Online Subset Selection over Matroids
arXiv 2609.24569 · 2026-09-23使用连接组约束的果蝇模型进行建筑图纸中一次性影线识别的主动视觉采样
What Survives on Real Drawings: Active Sampling, Connectome Wiring, and Matched Baselines in Architectural Document Vision
arXiv 2609.24565 · 2026-09-23MIRA：面向具身伴侣的实时全双工人机交互
MIRA: Real-Time Full-Duplex Human-Robot Interaction for Embodied Companions
arXiv 2609.24547 · 2026-09-23ME-VLM：用于具身认知与智能体协调的统一视觉语言模型
ME-VLM: A Unified VLM for Embodied Cognition and Agent Coordination
arXiv 2609.24526 · 2026-09-230.5%>100%：用于指代图像分割的双向互惠学习
0.5%>100%: Bidirectional Reciprocal Learning for Referring Image Segmentation
arXiv 2609.24510 · 2026-09-23混合精度DEIM-CUR分解的误差分析与精度选择
Error Analysis and Precision Selection for Mixed-Precision DEIM-CUR Decompositions
arXiv 2609.24509 · 2026-09-23VPRune：高效的无训练大语言模型前视觉令牌剪枝
VPRune: Efficient Training-free Pre-LLM Visual Token Pruning
arXiv 2609.24485 · 2026-09-23弱可见多边形中的见证集问题是多项式时间可解的
Witness Set in Weak Visibility Polygons is Polynomial-Time Solvable
arXiv 2609.24460 · 2026-09-23ActGov：通过策略约束验证治理LLM智能体行动
ActGov: Governing LLM Agent Actions via Policy-Constrained Validation
arXiv 2609.24446 · 2026-09-23基于Vision Transformer的相机空间手部姿态精确估计
Estimating Accurate Hand Pose in Camera Space with Vision Transformer
arXiv 2609.24424 · 2026-09-23Zeva-Ego：面向机器人操作的第一人称视角中训练与上下文因果学习
Zeva-Ego: Egocentric Mid-Training with In-Context Causal Learning for Robot Manipulation
arXiv 2609.24411 · 2026-09-23LiAuto-MindViT：一种具有自适应双向Mamba的混合视觉骨干网络
LiAuto-MindViT: A Hybrid Vision Backbone with Adaptive Bidirectional Mamba
arXiv 2609.24337 · 2026-09-23三视图几何中的三焦点张量
Trifocal Tensors in Three-View Geometry
arXiv 2609.24314 · 2026-09-23残差化何时有助于审计：格式效应、切片增益及其局限
When Residualization Helps an Audit: Format Effects, Slice Gains, and Their Limits
arXiv 2609.24194 · 2026-09-23StenoVLA-3D：面向胃肠道狭窄导航的3D感知推理VLA
StenoVLA-3D: 3D-Aware Reasoning VLA for Navigation Through Gastrointestinal Stenoses
arXiv 2609.24187 · 2026-09-23Re:CAP - 审计生产环境RAG管道中的检索覆盖率
Re:CAP - Auditing Retrieval Coverage in Production RAG Pipelines
arXiv 2609.24122 · 2026-09-23检验而非假定充分性：针对涌现网络结构校准生成式社会模拟器
Testing, not presuming, adequacy: calibrating generative social simulators against emergent network structure
arXiv 2609.24012 · 2026-09-23密度比重新评分用于不平衡分类
Density-Ratio Rescoring for Imbalanced Classification Using Raking Duals and Classifier Scores
arXiv 2609.23926 · 2026-09-23Majumdar--Papapetrou黑洞的精确视界响应与电荷系综
Exact horizon response and charge ensembles of Majumdar--Papapetrou black holes
arXiv 2609.23921 · 2026-09-23混合图模型的贝叶斯后验学习
Bayesian Posterior Learning of Mixed Graphical Models
arXiv 2609.23857 · 2026-09-23
