cs.LG（2026-06-25）

📊 共 31 篇论文

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (14) 支柱九：具身大模型 (Embodied Foundation Models) (14) 支柱八：物理动画 (Physics-based Animation) (1) 支柱一：机器人控制 (Robot Control) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (14 篇)

#题目一句话要点标签🔗⭐
1 A Generalization Theory for JEPA-Based World Models 提出JEPA基础世界模型的泛化理论以解决理论理解不足问题 world model world models JEPA
2 AIGP: An LLM-Based Framework for Long-Term Value Alignment in E-Commerce Pricing 提出AIGP框架以解决电商定价长期价值对齐问题 reinforcement learning offline reinforcement learning DPO
3 Sample-efficient Transfer Reinforcement Learning via Adaptive Reward Shaping and Policy-Ratio Reweighting Strategy 提出安全转移强化学习框架以解决高速公路换道决策中的训练不稳定问题 reinforcement learning policy learning reward shaping
4 Hallucination in World Models is Predictable and Preventable 提出MMBench2以解决世界模型中的幻觉问题 world model world models
5 State Representation Matters in Deep Reinforcement Learning: Application to Energy Trading 提出状态表示优化以提升深度强化学习在能源交易中的表现 reinforcement learning deep reinforcement learning
6 GEOALIGN: Geometric Rollout Curation for Robust LLM Reinforcement Learning 提出Geoalign以解决大语言模型强化学习中的方向不一致问题 reinforcement learning PPO large language model
7 Automating Potential-based Reward Shaping with Vision Language Model Guidance 提出VLM指导的潜在奖励塑形框架以解决稀疏奖励问题 reinforcement learning policy learning reward shaping
8 RolloutPipe: Overlapping Pipelined Rollout and Training in Disaggregated On-Policy LLM Reinforcement Learning 提出RolloutPipe以解决异构RLVR系统中的训练与回滚效率问题 reinforcement learning large language model
9 Reinforcement Learning without Ground-Truth Solutions can Improve LLMs 提出RiVER框架以解决无地面真实解的强化学习问题 reinforcement learning reward shaping
10 Sketched Linear Contrastive Learning: Approximation, Optimization, and Statistical Scaling 提出草图线性对比学习以解决对比学习的缩放规律问题 representation learning contrastive learning
11 Mean-Field PhiBE: Continuous-Time Mean-Field Reinforcement Learning from Discrete-Time Data 提出Mean-Field PhiBE以解决离散时间数据下的连续时间强化学习问题 reinforcement learning
12 Designing Reward Signals for Portable Query Generation: A Case Study in Industrial Semantic Job Search 提出RLAIF框架以优化工业语义职位搜索的查询生成 reinforcement learning reward shaping
13 Retroactive Advantage Correction: Closed-Form V-Trace Bias Correction for Delay-Aware RLHF 提出逆向优势修正以解决延迟感知的RLHF问题 reinforcement learning PPO RLHF
14 PEBS: Per-rater Empirical-Bayes Shrinkage for RLHF Reward-Model Calibration 提出PEBS以解决RLHF奖励模型校准问题 reinforcement learning RLHF

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
15 Can Large Language Models Reliably Code Qualitative Humanitarian Data? A Benchmark Study Against Human Expert Adjudication 评估大型语言模型在定性人道数据编码中的可靠性 large language model
16 When Does Quality-Aware Multimodal Fusion Matter? A Leakage-Safe Diagnostic for Decision-Level Dependence 提出泄漏安全诊断方法以评估多模态融合中的质量感知重要性 multimodal
17 A Causal Foundation Model for Structure and Outcome Prediction 提出TabPFN-CFM以解决因果结构与结果预测问题 foundation model
18 Just how sure are you? Improving Verbalized Uncertainty Calibration in Medical VQA 提出一种新方法以改善医学视觉问答中的不确定性校准 large language model multimodal
19 DMuon: Efficient Distributed Muon Training with Near-Adam Overhead 提出DMuon以解决分布式Muon优化器效率问题 large language model foundation model
20 When are likely answers right? On Sequence Probability and Correctness in LLMs 量化序列概率与正确性之间的关系以优化解码方法 large language model
21 Beyond the Hard Budget: Sparsity Regularizers for More Interpretable Top-k Sparse Autoencoders 提出稀疏正则化方法以提升Top-k稀疏自编码器的可解释性 foundation model
22 How Good Can Linear Models Be for Time-Series Forecasting? 提出基于线性模型的时间序列预测优化方法 foundation model
23 RSPC: A Benchmark for Modeling Stress and Psychiatric Conditions in Digitally Mediated Relationships using Psychiatrist Annotations 提出RSPC基准以解决数字化关系中的心理健康建模问题 large language model
24 PersistentKV: Page-Aware Decode Scheduling for Long-Context LLM Serving on Commodity GPUs 提出PersistentKV以解决长上下文LLM服务中的KV缓存问题 large language model
25 Zero-Shot Size Transfer for Neural ODEs on Sparse Random Graphs: Graphon Limits and Adjoint Convergence 提出零-shot大小转移方法以解决稀疏随机图上的神经ODE问题 zero-shot transfer
26 From Weights to Features: SAE-Guided Activation Regularization for LLM Continual Learning 提出SAE引导的激活正则化以解决LLM持续学习中的灾难性遗忘问题 large language model
27 Empirical Software Engineering TerraProbe: A Layered-Oracle Framework for Detecting Deceptive Fixes in LLM-Assisted Terraform 提出TerraProbe框架以检测LLM辅助Terraform中的欺骗性修复 large language model
28 Unified Zero-Shot Time Series Forecasting: A Darts Foundation 提出统一的零-shot时间序列预测框架以提升模型集成性 foundation model

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
29 fTNN: a tensor neural network for fractional PDEs 提出fTNN以解决分数阶偏微分方程问题 spatiotemporal

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
30 DroidBreaker: Practical and Functional Problem-Space Attacks on Machine-Learning Android Malware Detectors 提出DROIDBREAKER以解决现有Android恶意软件检测的攻击实用性问题 manipulation

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
31 Productionized Fairness Measurement Under Privacy Constraints 提出隐私保护的种族/民族公平性测量方法以解决隐私约束问题 OMOMO

⬅️ 返回 cs.LG 首页 · 🏠 返回主页