cs.LG（2025-04-04）

📊 共 19 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (9 🔗2) 支柱一：机器人控制 (Robot Control) (5) 支柱九：具身大模型 (Embodied Foundation Models) (5)

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#题目一句话要点标签🔗⭐
1 MORAL: A Multimodal Reinforcement Learning Framework for Decision Making in Autonomous Laboratories MORAL：用于自主实验室决策的多模态强化学习框架 reinforcement learning PPO embodied AI
2 Interpretable Multimodal Learning for Tumor Protein-Metal Binding: Progress, Challenges, and Perspectives 综述肿瘤蛋白-金属结合的可解释多模态学习，应对挑战并展望未来 predictive model multimodal
3 Decision SpikeFormer: Spike-Driven Transformer for Decision Making 提出Decision SpikeFormer，一种用于离线强化学习的脉冲驱动Transformer模型。 reinforcement learning offline RL offline reinforcement learning ✅
4 Enhanced Penalty-based Bidirectional Reinforcement Learning Algorithms 提出基于惩罚的双向强化学习算法，提升复杂环境下的策略学习能力 reinforcement learning policy learning
5 Autonomous state-space segmentation for Deep-RL sparse reward scenarios 提出基于自主状态空间分割的深度强化学习方法，解决稀疏奖励场景下的探索问题 reinforcement learning deep reinforcement learning policy learning
6 Improving Mixed-Criticality Scheduling with Reinforcement Learning 提出基于强化学习的混合关键性系统调度方法，提升任务完成率。 reinforcement learning
7 Generating ensembles of spatially-coherent in-situ forecasts using flow matching 提出基于流匹配的空间一致性集合预报方法，提升气象预测后处理性能。 flow matching
8 Optimizing Quantum Circuits via ZX Diagrams using Reinforcement Learning and Graph Neural Networks 提出基于ZX图、GNN和强化学习的量子电路优化方法，减少双量子比特门数量。 reinforcement learning
9 Semantic-guided Representation Learning for Multi-Label Recognition 提出语义引导的表征学习方法SigRL，解决多标签识别中语义信息不足的问题。 representation learning ✅

🔬 支柱一：机器人控制 (Robot Control) (5 篇)

#题目一句话要点标签🔗⭐
10 DML-RAM: Deep Multimodal Learning Framework for Robotic Arm Manipulation using Pre-trained Models DML-RAM：基于预训练模型的深度多模态学习机器人手臂操作框架 manipulation reinforcement learning multimodal
11 Offline and Distributional Reinforcement Learning for Wireless Communications 提出基于离线和分布强化学习的无线通信框架，解决6G网络中的不确定性和实时性挑战。 trajectory optimization reinforcement learning
12 Partially stochastic deep learning with uncertainty quantification for model predictive heating control 提出基于LSTM+BNN的部分随机深度学习模型，用于建筑供暖系统的模型预测控制。 MPC model predictive control predictive model
13 Recursive Training Loops in LLMs: How training data properties modulate distribution shift in generated data? 研究LLM递归训练中数据属性如何影响生成数据分布偏移 manipulation large language model
14 From Observation to Orientation: an Adaptive Integer Programming Approach to Intervention Design 提出一种自适应整数规划方法，用于在预算约束下优化因果干预设计。 manipulation

🔬 支柱九：具身大模型 (Embodied Foundation Models) (5 篇)

#题目一句话要点标签🔗⭐
15 Practical Poisoning Attacks against Retrieval-Augmented Generation 提出CorruptRAG，一种针对检索增强生成（RAG）系统的实用投毒攻击方法。 large language model
16 Identifying and Evaluating Inactive Heads in Pretrained LLMs 提出一种评估LLM中非活跃注意力头的方法，并通过消融实验验证其有效性。 large language model
17 Concept-based Rubrics Improve LLM Formative Assessment and Data Synthesis 基于概念的评分标准提升LLM在形成性评估和数据合成中的表现 large language model
18 HeterMoE: Efficient Training of Mixture-of-Experts Models on Heterogeneous GPUs HeterMoE：异构GPU上高效训练混合专家模型 large language model
19 Optimizing Specific and Shared Parameters for Efficient Parameter Tuning 提出SaS，通过优化特定和共享参数实现高效参数调优的PETL方法 foundation model

⬅️ 返回 cs.LG 首页 · 🏠 返回主页