2026-09论文中文摘要
Mira-Scene：用于生成式3D场景的像素对齐布局
Mira-Scene: Pixel-Aligned Layouts for Generative 3D Scene Reconstruction
arXiv 2609.23796 · 2026-09-23ScholarStack：面向科学智能体的分层研究资产编排与跨任务复用
ScholarStack: Layered Research Asset Orchestration and Cross-Task Reuse for Scientific Agents
arXiv 2609.23735 · 2026-09-23ELIPPS：部分自监督下逆问题的精确学习
ELIPPS: Exact Learning for Inverse Problems from Partial Self-supervision
arXiv 2609.23683 · 2026-09-23人类对齐的模型是人类的模型吗？偏好对齐中的图灵测试差距
Are Human-Aligned Models Models of Humans? A Turing-Test Gap in Preference Alignment
arXiv 2609.23640 · 2026-09-23RPMem：面向LLM智能体的跨会话长期循环参数化记忆学习
RPMem: Learning Long-Term Recurrent Parametric Memory Across Sessions for LLM Agents
arXiv 2609.23466 · 2026-09-23迈向可靠的水下潜水员-机器人交互：手势设计、交互逻辑与真实世界评估
Towards Reliable Underwater Diver-Robot Interaction: Gesture Design, Interaction Logic, and Real-World Evaluation
arXiv 2609.23392 · 2026-09-23为决策相关信念订立契约
Contracting for Decision-Relevant Beliefs
arXiv 2609.23285 · 2026-09-23CausalWM：具身世界模型的因果链式思维推理
CausalWM: Causal Chain-of-Thought Reasoning for Embodied World Model
arXiv 2609.23184 · 2026-09-23使用具有空间和时间借力的贝叶斯双广义贝塔回归对受限幸福感指数进行建模
Modeling bounded well-being indices using Bayesian double generalized beta regression with spatial and temporal borrowing
arXiv 2609.23120 · 2026-09-23248 keV LZ反冲：标量门户暗物质中非SM类夸克汤川耦合的可能迹象
The 248 keV LZ Recoil: A Possible Hint of Non-SM-Like Quark Yukawa Couplings with a Scalar-Portal Dark Matter
arXiv 2609.23096 · 2026-09-23离散与连续：LALMs中统一音频理解的综合研究
Discrete vs. Continuous: A Comprehensive Study of Unified Audio Understanding in LALMs
arXiv 2609.22851 · 2026-09-23全身UMI：通过实时运动生成将UMI操作技能迁移至人形全身操作
Whole-Body UMI: Transferring UMI Manipulation Skills to Humanoid Whole-Body Manipulation via Real-Time Motion Generation
arXiv 2609.22829 · 2026-09-23从IceCube到IT-Sphere：用于银行IT根因分析的混合量子-经典GNN
From IceCube to IT-Sphere: A Hybrid Quantum-Classical GNN for Banking IT Root Cause Analysis
arXiv 2609.22822 · 2026-09-23超越平均误差：面向时间序列预测的预言机知情压力测试
Beyond Average Error through Oracle-Informed Stress Tests for Time-Series Forecasting
arXiv 2609.22820 · 2026-09-23参数化稠密-稀疏融合用于混合检索：在Qdrant上对BEIR SciFact的排名-分数混合进行调优
Parameterized Dense-Sparse Fusion for Hybrid Retrieval: Tuning a Rank-Score Mix on BEIR SciFact with Qdrant
arXiv 2609.22770 · 2026-09-23$f$-形变形式下的自旋-玻色子映射
Spin-Boson Mappings in the Formalism of $f$-Deformations
arXiv 2609.22502 · 2026-09-23磁化主序星演化形成的超钱德拉塞卡白矮星：来自STARS模拟的新质量极限
Super-Chandrasekhar white dwarfs by the evolution of magnetized main-sequence stars: New mass-limits from STARS simulation
arXiv 2609.22437 · 2026-09-23差分隐私与公平性审计的不规则纵向健康记录分数扩散
Differentially Private and Fairness-Audited Score Diffusion for Irregular Longitudinal Health Records
arXiv 2609.22401 · 2026-09-23一种从低动量转移电子散射确定208Pb电荷密度第4阶矩的新方法
A novel method to determine the 4th moment of the 208Pb charge density from low-momentum-transfer electron scattering
arXiv 2609.22387 · 2026-09-23将大规模邻域搜索的破坏步骤重塑为密集分割
Recasting the Destroy Step of Large Neighborhood Search as Dense Segmentation
arXiv 2609.22350 · 2026-09-23AffordanceWAM：面向机器人操作的可供性感知联合世界-动作建模
AffordanceWAM: Affordance-Aware Joint World-Action Modeling for Robot Manipulation
arXiv 2609.22332 · 2026-09-23第五Bessel矩的一个二次临界值猜想
A quadratic critical-value conjecture for the fifth Bessel moment
arXiv 2609.22318 · 2026-09-23测试大脑和语言模型中的两态访问：人类EEG复现、模拟审计及提出的模型检测方法
The Cross-Substrate Access Assay: What an Indicator Test Must Declare to Travel from Brain to Language Model
arXiv 2609.22300 · 2026-09-23事件流时间量化中的鲁棒性-分辨率权衡
On The Robustness-Resolution Tradeoff In Temporal Quantization Of Event Streams
arXiv 2609.22295 · 2026-09-23超越任务完成：训练有能力且安全的计算机使用智能体
Beyond Task Completion: Training Capable and Safe Computer-Use Agents
arXiv 2609.22178 · 2026-09-23StepKV：面向LLM智能体的步骤感知KV缓存压缩
StepKV: Step-Aware KV Cache Compression for LLM Agents
arXiv 2609.22158 · 2026-09-23PAGE: 分区感知的门控KV缓存驱逐
PAGE: Partition-Aware Gated KV-Cache Eviction
arXiv 2609.22157 · 2026-09-23$λ$-Controlled GRPO：将流匹配比率不稳定性转化为预算化资源
$λ$-Controlled GRPO: Turning Flow-Matching Ratio Instability into a Budgeted Resource
arXiv 2609.22041 · 2026-09-23可编程衍射神经网络的光学模式排序
Optical Mode Sorting with a Programmable Diffractive Neural Network
arXiv 2609.22015 · 2026-09-23贝叶斯信念层：用于可控LLM智能体意见动态的贝叶斯编年史智能体
Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents
arXiv 2609.21997 · 2026-09-23基于部分充电数据的锂离子电池剩余使用寿命联合预测与容量估计
Joint Remaining Useful Life Prediction and Capacity Estimation of Lithium-Ion Batteries Using Partial-Charging Data
arXiv 2609.21932 · 2026-09-23蓝鸟和模仿鸟不能产生不动点组合子
Bluebirds and mockingbirds cannot produce a fixed-point combinator
arXiv 2609.21765 · 2026-09-23加权Kolmogorov方程：传递估计、局部有界性与Harnack不等式
Weighted Kolmogorov equations: transfer estimates, local boundedness, and Harnack inequalities
arXiv 2609.21737 · 2026-09-23ZYT-World：用于闭环自动驾驶仿真的实时可控世界模型
ZYT-World: A Real-Time Controllable World Model for Closed-Loop Autonomous-Driving Simulation
arXiv 2609.21712 · 2026-09-23巨原子诱导耦合腔阵列中的高阶输出零点
Giant-Atom-Induced High-Order Output Zeros in a Coupled-Cavity Array
arXiv 2609.21560 · 2026-09-23Chern-Simons函数与量子霍尔效应
The Chern-Simons Function and the Quantum Hall Effect
arXiv 2609.21182 · 2026-09-23人类神经元放电的磁场：能否通过量子传感MRI测量？
Magnetic Field of Firing Neuron in Humans: Measurable by Quantum Sensing MRI?
arXiv 2609.21106 · 2026-09-23竞争性地理解释的早期排序的生成性反演
Ranking Competing geologic interpretations via foundation-model-assisted generative hydrologic inversion
arXiv 2609.20978 · 2026-09-23不要责怪模型，验证数据：基于SMT的数据集验证评估
Don't Blame the Model, Verify the Data: An Evaluation of SMT-based Dataset Verification (Extended Version)
arXiv 2609.20959 · 2026-09-23白矮星诞生踢作为速度空间随机游走. I. Fokker-Planck理论
White Dwarf Natal Kicks as Velocity-Space Random Walks. I. Fokker-Planck Theory
arXiv 2609.20940 · 2026-09-23连续延迟记忆随机梯度下降与基于天体物理时间序列历史的连续时间强化学习
Continuous Delayed-Memory Stochastic Gradient Descent and Continuous-Time Reinforcement Learning from History of Astrophysical Time Series Studies
arXiv 2609.20906 · 2026-09-23稳定子公钥认证：部分预测界与密钥重用的极限
Stabilizer-Public-Key Authentication: Long Stabilizer Public Keys Resist Finite-Copy Forgery Attacks
arXiv 2609.20877 · 2026-09-23MME-Safety：多模态大语言模型安全性的细粒度基准
MME-Safety: A Fine-grained Benchmark for Safety Evaluation of MLLMs
arXiv 2609.20850 · 2026-09-23量化前沿大语言模型智能体的过度声称倾向
Quantifying Overclaiming Propensity in Frontier LLM Agents
arXiv 2609.20812 · 2026-09-23DocAttriBench：文档视觉问答中的答案溯源基准
DocAttriBench: Benchmarking Answer Grounding in Document Visual Question Answering
arXiv 2609.20574 · 2026-09-23平均曲率流完备自收缩子的刚性定理
The rigidity theorems for complete self-expander of mean curvature flow
arXiv 2609.20429 · 2026-09-23标记关联结构：用于文本、知识图谱与超图的原生Transformer建模
Labeled Incidence Structures for Native Transformer Modeling of Text, Knowledge Graphs, and Hypergraphs
arXiv 2609.20278 · 2026-09-23无表索引的锥形记忆化网格：排序参数函数的紧凑外核评估
Memoization Without Keys: Compact, Out-of-Core Tables for Functions of Sorted Arguments
arXiv 2609.20276 · 2026-09-23矩阵分析中若干开放问题的解
Solutions to some open problems in matrix analysis
arXiv 2609.20094 · 2026-09-23群对称下的加权对偶闵可夫斯基问题
The Weighted Dual Minkowski Problem Under Group Symmetry
arXiv 2609.20047 · 2026-09-23
