2026-10论文中文摘要
语言条件下的标记与推理效率在大语言模型中的研究：一项配对跨语言研究协议
Language-Conditioned Token and Reasoning Efficiency in Large Language Models: A Paired Cross-Lingual Study Protocol
arXiv 2610.04295 · 2026-10-06利用机器学习获取320万条DESI DR1光谱的恒星参数与丰度
Stellar parameters and abundances for 3.2 million DESI DR1 spectra using machine learning
arXiv 2610.04294 · 2026-10-06超外围碰撞中独占 $J/\psi+\gamma$ 产生的振幅级次 NLO 预言
Amplitude-level NLO predictions for exclusive $J/ψ+γ$ production in ultraperipheral collisions
arXiv 2610.04293 · 2026-10-06LMBuild：评估LLM智能体生成可构建且功能完备的结构
LMBuild: Evaluating LLM Agents for Generating Buildable and Functional Structures
arXiv 2610.04292 · 2026-10-06局部Fisher信息实现稀疏因果发现
Local Fisher Information Enables Sparse Causal Discovery
arXiv 2610.04291 · 2026-10-06三元锑化物La$T$Sb$_2$ ($T$ = Cu, Ag)的磁输运与电子结构研究
Magneto-transport and electronic structure studies of ternary antimonides, La$T$Sb$_2$ ($T$ = Cu, Ag)
arXiv 2610.04290 · 2026-10-06无线基础模型的缩放：基于表示多样性与多分支架构
Scaling of Wireless Foundation Models via Representation Diversity and Multi-Branch Architectures
arXiv 2610.04289 · 2026-10-06最大纠缠态判别中单向与双向LOCC之间的多项式分离
Polynomial Separation between One-Way and Two-Way LOCC in Discrimination of Maximally Entangled States
arXiv 2610.04288 · 2026-10-06VIGIL：用于电子表格问答的验证器知情门控改进循环
VIGIL: Verifier-Informed Gated Improvement Loop for Spreadsheet Question Answering
arXiv 2610.04287 · 2026-10-06CyTReX：基于可解释人工智能的DER网络安全威胁推理框架
CyTReX: Explainable AI-Based Cybersecurity Threat Reasoning Framework for DER Networks
arXiv 2610.04286 · 2026-10-06摊销得分-哈密顿策略迭代：松弛随机控制问题的无网格方案
Amortized Score-Hamiltonian Policy Iteration: A Grid-Free Scheme for Relaxed Stochastic Control Problems
arXiv 2610.04285 · 2026-10-06超越尘埃的局部宇宙学：协变动力学理论
Beyond dust in local cosmography: covariant kinetic theory
arXiv 2610.04284 · 2026-10-06一阶引导：将权重适应转化为激活引导
First-Order Steering: Translating Weight Adaptation into Activation Steering
arXiv 2610.04283 · 2026-10-06中子星中的暗物质：全广义相对论下的双流体 $f$-模振荡
Dark matter in neutron stars: two-fluid $f$-mode oscillations in full general relativity
arXiv 2610.04282 · 2026-10-06OctMesh：一种用于无损三角网格压缩的统一八叉树层次框架
OctMesh: A Unified Octree-Hierarchical Framework for Lossless Triangle Mesh Compression
arXiv 2610.04281 · 2026-10-06统一几何状态中的密集神经符号推理
Dense Neuro-Symbolic Reasoning in a Unified Geometry State
arXiv 2610.04280 · 2026-10-06计算带共享约束的整数规划博弈中的均衡
Computing Equilibria in Integer Programming Games with Shared Constraints
arXiv 2610.04279 · 2026-10-06RUL 解释是否站得住脚？C-MAPSS 上归因的忠实性与稳定性
Do RUL explanations hold up? Faithfulness and stability of attributions on C-MAPSS
arXiv 2610.04278 · 2026-10-06SyclKittens：面向Intel GPU的程序员与编码代理的瓦片编程模型
SyclKittens: A Tile Programming Model for Programmers and Coding Agents on Intel GPUs
arXiv 2610.04277 · 2026-10-06当隐蔽性需要记忆：白噪声约束下的预算攻击调度
When Stealth Requires Memory: Budgeted Attack Scheduling under a Whiteness Constraint
arXiv 2610.04276 · 2026-10-06符号搜索并未穷尽：Lean4 中持久证明空间探索
Symbolic Search Is Not Exhausted: Persistent Proof-Space Exploration in Lean4
arXiv 2610.04275 · 2026-10-06具有给定模数奇素数除数的类换位群
Groups of Class Transpositions with Prescribed Prime Divisors of the Moduli
arXiv 2610.04274 · 2026-10-06集成插补-分类用于缺失数据的监督学习
Integrated Imputation-Classification for Supervised Learning with Missing Data
arXiv 2610.04273 · 2026-10-06重新思考用于多教师能力融合的自蒸馏
Rethinking Self-Distillation for Multi-Teacher Capability Merging
arXiv 2610.04272 · 2026-10-06具有近指数 sofic 轮廓的可解群，以及需要近线性内存的量子信道
Amenable groups with nearly exponential sofic profile, and quantum channels that need nearly linear memory
arXiv 2610.04271 · 2026-10-06向量空间的完全非平凡交定理
The complete nontrivial-intersection theorem for vector spaces
arXiv 2610.04270 · 2026-10-06自我反思微调：从失败经验中增强智能体对提示注入攻击的安全性
Self-Reflection Fine-Tuning: Enhancing Agent Security against Prompt Injection Attacks from Failure Experience
arXiv 2610.04269 · 2026-10-06漂移势的逆随机控制Galerkin估计器
An inverse-stochastic-control Galerkin estimator for drift potentials
arXiv 2610.04268 · 2026-10-06AI赋能的单选题质量保障
AI-Enabled Quality Assurance for Multiple-Choice Assessment Items
arXiv 2610.04267 · 2026-10-06偏移冗余机械臂的实时共形播种混合逆运动学
Real-Time Conformal-Seeded Hybrid Inverse Kinematics for Offset Redundant Manipulators
arXiv 2610.04266 · 2026-10-06来自热阈值的赝南布-戈德斯通玻色子运动
Pseudo-Nambu-Goldstone Boson Motion from Thermal Thresholds
arXiv 2610.04265 · 2026-10-06自适应Bregman交替投影用于可行Gromov-Wasserstein学习
Adaptive Bregman Alternating Projections for Feasible Gromov-Wasserstein Learning
arXiv 2610.04264 · 2026-10-06多模态双编码器检索用于自动ICD编码
Multimodal Dual-Encoder Retrieval for Automated ICD Coding
arXiv 2610.04263 · 2026-10-06CADForge：基于显式几何推理的智能体式单视图CAD重建
CADForge: Agentic Single-View CAD Reconstruction with Explicit Geometry Reasoning
arXiv 2610.04262 · 2026-10-06使用强化微调大型语言模型进行社交推理游戏
Playing social deduction games with reinforcement fine-tuned large language models
arXiv 2610.04261 · 2026-10-06潜意识学习中的跨特质迁移
Cross-Trait Transfer in Subliminal Learning
arXiv 2610.04260 · 2026-10-06非局部非凸泛函族的尖锐Poincaré不等式
Sharp Poincaré Inequalities for Families of Nonlocal Nonconvex Functionals
arXiv 2610.04259 · 2026-10-06FloVMos：基于光流的医学视频拼接
FloVMos: Optical Flow-based Medical Video Mosaicking
arXiv 2610.04258 · 2026-10-06一维完全可压缩Navier--Stokes--Poisson方程激波剖面的存在性与稳定性
Existence and stability of shock profiles for a one-dimensional full compressible Navier--Stokes--Poisson equations
arXiv 2610.04257 · 2026-10-06一个可手工验证的证明：两个隐藏 ReLU 层可计算六个数的最大值
A Hand-Checkable Proof That Two Hidden ReLU Layers Compute the Maximum of Six Numbers
arXiv 2610.04256 · 2026-10-06扎根于时间：机器人操作中时间定位的多源数据集与基准
Grounded in Time: A Multi-Source Dataset and Benchmark for Temporal Grounding in Robotic Manipulation
arXiv 2610.04255 · 2026-10-06Ennola三次域中的指定理想类与椭圆点
Prescribed ideal classes and elliptic points in Ennola cubic fields
arXiv 2610.04254 · 2026-10-06Spec2Game：大型语言模型能否根据详细规格生成完整可玩的游戏？
Spec2Game: Can LLMs Generate Complete Playable Games from Detailed Specifications?
arXiv 2610.04253 · 2026-10-06从墙结构到闭镜像对称性：$K_{\mathbb{P}^2}$ 的情形——正实轨迹上的重整化周期、来自墙函数的闭 Gromov-Witten 不变量以及热带枚举
From wall structures to closed mirror symmetry. The case of $K_{\mathbb{P}^2}$: Renormalized periods over the positive real locus, closed Gromov-Witten invariants from wall functions, and tropical enumeration
arXiv 2610.04252 · 2026-10-06高维中柯西组合 Hettmansperger-Randles 位置检验
Cauchy-Combined Hettmansperger-Randles Location Tests in High Dimensions
arXiv 2610.04251 · 2026-10-06ROOT：为用户指定的具身行为发现奖励函数
ROOT: Discovering Rewards for User-Specified Embodied Behaviors
arXiv 2610.04250 · 2026-10-06优化器几何设定节奏：矩阵分解中的谱学习动力学
Optimizer Geometry Sets the Pace: Spectral Learning Dynamics in Matrix Factorization
arXiv 2610.04249 · 2026-10-06面外类阻尼自旋轨道矩的材料与计量学
Materials and Metrology for Out-of-Plane Damping-Like Spin-Orbit Torque
arXiv 2610.04248 · 2026-10-06Q-MERGE：面向大规模经典数据的量子态制备并行化
Q-MERGE: Parallelising Quantum State Preparation for Large-Scale Classical Data
arXiv 2610.04247 · 2026-10-06生成式智能体中的心理动力学基准测试
Benchmarking Psychological Dynamics in Generative Agents
arXiv 2610.04246 · 2026-10-06
