2026-10论文中文摘要
定价引导：语言模型能否生成未来研究想法？
Priced Guidance: Can Language Models Generate Future Research Ideas?
arXiv 2610.04976 · 2026-10-06长时程工具使用AI智能体中自生成子目标的运行时授权
Runtime Authorization of Self-Generated Subgoals in Long-Horizon Tool-Using AI Agents
arXiv 2610.04975 · 2026-10-06IREA：基于中间表示的嵌入对齐用于规范性RAG
IREA: Intermediate Representation-based Embedding Alignment for Normative RAG
arXiv 2610.04974 · 2026-10-06TrajLong：为中期训练协同设计智能体与长上下文监督
TrajLong: Co-Designing Agentic and Long-Context Supervision for Mid-Training
arXiv 2610.04973 · 2026-10-06Riemann xi-函数泰勒系数的无限对数凹性
Infinite log-concavity of the Taylor coefficients of the Riemann xi-function
arXiv 2610.04972 · 2026-10-06探测HL-LHC顶夸克对产生中bW自旋关联的反常Wtb耦合
Probing anomalous $Wtb$ couplings through $bW$ spin correlations in top-quark pair production at the HL-LHC
arXiv 2610.04971 · 2026-10-06整合组学揭示美国牛奶生物活性变异的可操作驱动因素
Integrated omics reveals actionable drivers of bioactive variation in US milk
arXiv 2610.04970 · 2026-10-06哈密顿度量学习与基于能量的训练：一种用于优化的耗散几何框架
Hamiltonian Metric Learning and Energy-Based Training: A Dissipative Geometric Framework for Optimization
arXiv 2610.04969 · 2026-10-06强 $\ell^r$ 球面极大函数
The strong $\ell^r$ spherical maximal function
arXiv 2610.04968 · 2026-10-06一个 Token 就足够：用前缀引导桥接提示与激活引导
One Token Can Be Enough: Bridging Prompting and Activation Steering with Prefix Steering
arXiv 2610.04967 · 2026-10-06BACAM：面向多轮交互的行为感知持续智能体合并
BACAM: Behavior-Aware Continual Agent Merging for Multi-Turn Interaction
arXiv 2610.04966 · 2026-10-06计数环面三维簇上的高阶层
Counting higher-rank sheaves on toric threefolds
arXiv 2610.04965 · 2026-10-06一种用于拟合具有过度离散性的有限离散数据的新型Bernstein-二项模型：基于似然和贝叶斯方法
A new Bernstein-binomial model for fitting finite discrete data with over-dispersion: Likelihood-based and Bayesian approaches
arXiv 2610.04964 · 2026-10-06MAGIC：拓扑感知的解析式图小样本类增量学习
MAGIC: Topology-Aware Analytic Graph Few-Shot Class-Incremental Learning
arXiv 2610.04963 · 2026-10-06$BF$ 有效作用的拓扑意义
$BF$ effective action topological implications
arXiv 2610.04962 · 2026-10-06以深度学习方式构建LLM智能体系统：从模块化设计到架构搜索
Building LLM Agent Systems the Deep Learning Way: From Modular Design to Architecture Search
arXiv 2610.04961 · 2026-10-06大维度离散自相似坍缩中的Polyakov反作用
Polyakov backreaction in large D discretely self-similar collapse
arXiv 2610.04960 · 2026-10-062H-NbSe2中体态金属能带的Floquet修饰
Floquet dressing of bulk metallic bands in 2H-NbSe2
arXiv 2610.04958 · 2026-10-06Trinity：一个用于训练、精炼和评分生成式布局规划器的可微分物理模型
Trinity: One Differentiable Physics for Training, Refining and Scoring Generative Floorplanners
arXiv 2610.04957 · 2026-10-06从过载到有保障：面向LoRA辅助本地LLM部署的高吞吐多SLO执行
From Overloaded to Guaranteed: High-Throughput Multi-SLO Enforcement for LoRA-Assisted On-Premise LLM Deployment
arXiv 2610.04956 · 2026-10-06Kapture：利用Koopman支配学习捕获心脏动力学，实现基于雷达的高效心电恢复
Kapture: Capturing Cardiac Dynamics with Koopman-Governed Learning for Efficient Radar-Based Electrocardiogram Recovery
arXiv 2610.04955 · 2026-10-06探索未来质子-质子对撞机与超神冈实验中的规范传递超对称大统一理论
Exploring the Supersymmetric Grand Unified Theories with Gauge Mediation at the Future Proton-Proton Colliders and Hyper-Kamiokande Experiment
arXiv 2610.04954 · 2026-10-06EDISCO：用于欧几里得组合优化的等变离散扩散模型
EDISCO: Equivariant DIScrete Diffusion for Euclidean Combinatorial Optimization
arXiv 2610.04953 · 2026-10-06局部哈密顿量的带组合可靠性的间隙放大
Gap Amplification for Local Hamiltonians with Combinatorial Soundness
arXiv 2610.04952 · 2026-10-06LLBPE：基于链表的GPU并行BPE分词器
LLBPE: Linked-List Based GPU-Parallel BPE Tokenizer
arXiv 2610.04951 · 2026-10-06教师应如何准备？基于学生诱导状态的强化学习用于在线策略蒸馏
How Should Teachers Be Prepared? RL on Student-Induced States for On-Policy Distillation
arXiv 2610.04950 · 2026-10-06统一动力学框架：NVIDIA Isaac Sim中六自由度管道跟踪ROV的强化学习与经典控制
A Unified Dynamics Framework for Reinforcement Learning and Classical Control of a Six-DOF Pipeline-Tracking ROV in NVIDIA Isaac Sim
arXiv 2610.04949 · 2026-10-06社会最优性并不意味着无权Max-k-Cut博弈中的联盟稳定性
Social Optimality Does Not Imply Coalitional Stability in Unweighted Max-k-Cut Games
arXiv 2610.04948 · 2026-10-06非正则图的渐近谱半径
Asymptotic spectral radius of nonregular graphs
arXiv 2610.04947 · 2026-10-06TempoBridge：基于最优传输耦合的源条件流匹配用于单细胞群体转变
TempoBridge: Source-Conditioned Flow Matching with Optimal Transport Couplings for Single-Cell Population Transitions
arXiv 2610.04945 · 2026-10-06有限和非凸-强凹极小极大优化的近最优复杂度
Near-Optimal Complexity of Finite-Sum Nonconvex-Strongly-Concave Minimax Optimization
arXiv 2610.04944 · 2026-10-06透过Eisenstein理想的视角
Through the lens of the Eisenstein ideal
arXiv 2610.04943 · 2026-10-06零样本时间序列问答：通过解耦感知与推理
Zero-Shot Time-Series Question Answering via Decoupled Perception and Reasoning
arXiv 2610.04942 · 2026-10-06由热核协方差的高斯噪声驱动的随机热方程的中央极限定理
Central limit theorems for stochastic heat equation driven by Gaussian noise with heat-kernel covariance
arXiv 2610.04941 · 2026-10-06软件世界模型：从后果预测到决策价值
Software World Models: From Consequence Prediction to Decision Value
arXiv 2610.04940 · 2026-10-06基于图的检查与干预工具：评估PINN中的机制学习
A Graph-Based Inspection and Intervention Tool for Assessing Mechanistic Learning in PINNs
arXiv 2610.04939 · 2026-10-06D-DOIT：通过Doob's h-变换实现离散扩散的无训练适应
D-DOIT: Training-free Adaptation of Discrete Diffusion via Doob's h-Transform
arXiv 2610.04938 · 2026-10-06在部分可观测动态博弈中针对未知对手编排K级策略
Orchestrating Level-$K$ Policies Against Unknown Opponents in Partially-Observable Dynamic Games
arXiv 2610.04937 · 2026-10-06局部少数类不平衡下的学习
Learning under Localized Minority Imbalance
arXiv 2610.04936 · 2026-10-06添加盐对单价金属离子和模型水溶性聚合物界面动力学的影响
Effect of Added Salts on the Interfacial Dynamics of Monovalent Metal Ions and Model Water-Soluble Polymers
arXiv 2610.04934 · 2026-10-06DiVeR：用于VLA测试时扩展的决策关键验证器学习
DiVeR: Decision-Critical Verifier Learning for VLA Test-Time Scaling
arXiv 2610.04933 · 2026-10-06温度和添加盐对稀溶液中模型聚两性电解质聚合物的影响
Effect of Temperature and Added Salt on a Model Polyzwitterion Polymer in Dilute Solution
arXiv 2610.04932 · 2026-10-06十亿级未策划短视频缩略图优化：基于多臂老虎机方法
Billion-Scale Thumbnail Optimization for Uncurated Short-Form Videos via Multi-Armed Bandits
arXiv 2610.04931 · 2026-10-06成对检测、手性转换：量子多态的资源理论
Pairwise detection, chiral conversion: Resource theory of quantum multistates
arXiv 2610.04930 · 2026-10-06提示主导性与不对称验证器成本：1B规模下GRPO在GSM8K上的经验消融研究
Prompt Dominance and Asymmetric Verifier Costs: Empirical Ablations of GRPO at 1B Scale on GSM8K
arXiv 2610.04928 · 2026-10-06为智能体机器学习工程系统汇聚洞察
Assembling Insights for Agentic Machine Learning Engineering Systems
arXiv 2610.04927 · 2026-10-06机器学习模型用于慢性肾脏病分类的可复现性与泄漏控制评估
Reproducibility and Leakage-Controlled Evaluation of Machine-Learning Models for Chronic Kidney Disease Classification
arXiv 2610.04926 · 2026-10-06TSAE：用于解释时间序列预测模型的结构化稀疏自编码器
TSAE: Structured Sparse Autoencoders for Interpreting Time-Series Forecasting Models
arXiv 2610.04925 · 2026-10-06先轻推再推动：基于触觉探测的物理感知导航
Nudge Before You Push: Physics-Aware Navigation via Tactile Probing
arXiv 2610.04924 · 2026-10-06提升智能家居智能体的韧性
Increasing Resilience of Smart Home Agents
arXiv 2610.04923 · 2026-10-06
