cs.LG（2026-06-22）

📊 共 34 篇论文 | 🔗 3 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (15 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (15 🔗1) 支柱一：机器人控制 (Robot Control) (3) 支柱四：生成式动作 (Generative Motion) (1 🔗1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (15 篇)

#题目一句话要点标签🔗⭐
1 Dynamic multi-agent deep reinforcement learning-based pricing and incentivization approach in multimodal transportation networks 提出基于动态多智能体深度强化学习的定价与激励方法以解决多模式交通网络问题 reinforcement learning deep reinforcement learning multimodal
2 LOLLA: Deep Reinforcement Learning for Closed-Loop Link Adaptation Towards a GPU-Accelerated AI-RAN 提出LOLLA以解决高动态环境下的链路适应问题 reinforcement learning deep reinforcement learning PPO
3 PG-MAP: Joint MAP Optimization for Inference-Time Alignment of Diffusion and Flow-Matching Models 提出PG-MAP以解决文本到图像模型推理对齐问题 flow matching classifier-free guidance latent optimization
4 Patient-Aware Contrastive Learning Preserves Per-Patient Structure in RR-Interval Representations 提出患者感知对比学习以解决RR间期表示中的个体结构保留问题 representation learning contrastive learning geometric consistency
5 Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learning 提出G2PO以解决长时间跨度强化学习中的奖励稀疏问题 reinforcement learning large language model
6 Provable Benefits of RLVR over SFT for Reasoning Models: Learning to Backtrack Efficiently 提出RLVR方法以提升推理模型的回溯效率 reinforcement learning large language model chain-of-thought
7 MORL-A2C: Multi-Objective Reinforcement Learning Reranker for Optimizing Healthiness in MOPI-HFRS 提出MORL-A2C以优化MOPI-HFRS中的健康饮食推荐 reinforcement learning behavior cloning
8 Superhuman AI for Generals.io Using Self-Play Reinforcement Learning 提出超人类AI以解决Generals.io游戏中的策略挑战 reinforcement learning transformer policy
9 Self-Evolution for Multi-Turn Tool-Calling Agents via Divergence-Point Preference Learning 提出自我进化方法以提升多轮工具调用代理的性能 preference learning DPO
10 ReNIO: Reweighting Negative Trajectory Importance for LLM On-Policy Distillation 提出ReNIO以解决LLM蒸馏中负向轨迹重要性的问题 reinforcement learning distillation ✅
11 One-Step Flow Matching for Generative Modeling of Path-Dependent Physical Fields 提出基于流匹配的生成模型以高效模拟路径依赖物理场 flow matching distillation
12 Bridge the Gaps: Heterogeneous Attributed Graph Clustering via Quaternion Representation Learning 提出AGREE框架以解决异构属性图聚类问题 representation learning
13 EvoRubrics: Dynamic Rubrics as Rewards via Adversarial Co-Evolution for LLM Reinforcement Learning 提出EvoRubrics以解决静态评分标准在强化学习中的不足问题 reinforcement learning
14 TaLK: Text-attributed Graph Dataset Distillation via Coupling Language Model with Graph-Aware Kernel 提出TaLK以解决文本属性图数据蒸馏问题 distillation
15 Hierarchical Reinforcement Learning for Sparse-Reward Search in Commutative Algebra 提出层次强化学习框架以解决稀疏奖励的代数问题 reinforcement learning

🔬 支柱九：具身大模型 (Embodied Foundation Models) (15 篇)

#题目一句话要点标签🔗⭐
16 Physiology-Aware CNN and Zero-Shot Multimodal LLMs for ECG Image Classification: A Comparative Study 提出生理感知CNN与零-shot多模态LLMs以解决ECG图像分类问题 large language model multimodal
17 SVD-Surgeon: Optimal Singular-Value Surgery for Large Language Model Compression 提出SVD-Surgeon以解决大语言模型压缩问题 large language model
18 When AUC 0.998 Is Not Enough: A Candidate Evaluation Protocol for Hidden-State Probes of Indirect Prompt Injection in Multimodal Computer-Use Agents 提出隐状态探测协议以评估多模态计算代理中的间接提示注入问题 multimodal
19 Retrieval-Augmented Multimodal Learning for Enzyme-Substrate Interaction Prediction Under Low-Homology Shift 提出RAMMESI框架以解决酶-底物相互作用预测中的低同源性问题 multimodal
20 On the Limits of Prompt-Conditioned Language Models as General-Purpose Learners 提出语言模型的局限性分析以解决通用学习问题 large language model multimodal
21 Distribution-Aware Diffusion-LLM for Robust Ultra-Long-Term Time Series Forecasting 提出Diffusion-LLM以解决超长时间序列预测中的多模态挑战 large language model multimodal
22 Open Problem: Is AdamW Effective Under Heavy-Tailed Noise? 探讨AdamW在重尾噪声下的有效性问题 large language model
23 Physics-Informed Modeling for Wood Thermal Analysis and Prediction 提出物理信息建模以解决木材热分析与预测问题 multimodal ✅
24 FlexServe: A Fast and Secure LLM Serving System for Mobile Devices with Flexible Resource Isolation 提出FlexServe以解决移动设备LLM推理的安全与效率问题 large language model
25 GRIMIP: A General Framework for Instance-Specific Configuration of MIP Solvers Using LLMs 提出GRIMIP以解决MIP求解器超参数配置问题 large language model
26 Memory Contagion: Cross-Temporal Propagation of Evaluator Bias via Agent Memory 提出记忆传播机制以解决评估者偏见问题 large language model
27 Scalable Physics-Inspired Transformers for Spin Glasses 提出物理启发的变压器以解决自旋玻璃的高效采样问题 large language model
28 EEG Benchmarking Needs a Task Specification Layer: NeuroDoc for Rulebook-Guided, Executable Benchmark Construction 提出NeuroDoc以解决EEG基准测试任务规范化问题 foundation model
29 Priority-Aware Learning-Unlearning Correction for Dynamic Decentralized LoRA Fine-Tuning 提出优先级感知学习-遗忘校正框架以解决动态去中心化LoRA微调问题 large language model
30 Noise is Signal: Density-Based Outliers as Leading Indicators of Occupational Emergence in Labor Market Text 提出噪声信号理论以识别劳动市场新兴职业 foundation model

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
31 Attacking the Trusted Imagination: Oracle-Level Integrity Attacks on Imagine-then-Act World Models 提出针对想象-行动模型的完整性攻击方法以增强安全性 MPC model predictive control world model
32 Learning Process Rewards via Success Visitation Matching for Efficient RL 提出通过成功访问匹配学习过程奖励以解决稀疏奖励问题 manipulation reinforcement learning
33 The Unseen Hand: Manipulating Model Fairness and SHAP with Targeted Identity Re-Association Attacks 提出目标身份重关联攻击以操控模型公平性和SHAP解释 manipulation

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
34 Scheduling Thoughts: Learning the Order of Thought in Diffusion Language Models 提出自我感知调度以优化扩散语言模型的思维顺序 MDM ✅

⬅️ 返回 cs.LG 首页 · 🏠 返回主页