cs.LG（2025-11-28）

📊 共 21 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (9 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗2) 支柱一：机器人控制 (Robot Control) (1) 支柱四：生成式动作 (Generative Motion) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (9 篇)

#题目一句话要点标签🔗⭐
1 Quantized-Tinyllava: a new multimodal foundation model enables efficient split learning 提出Quantized-Tinyllava，通过量化压缩实现高效的联邦学习多模态大模型训练。 foundation model multimodal
2 Transformer-Driven Triple Fusion Framework for Enhanced Multimodal Author Intent Classification in Low-Resource Bangla 提出BangACMM框架，通过Transformer驱动的三重融合提升低资源孟加拉语作者意图分类。 multimodal
3 EEG-Bench: A Benchmark for EEG Foundation Models in Clinical Applications EEG-Bench：用于评估脑电图临床应用基础模型的统一基准测试框架 foundation model
4 LLM4XCE: Large Language Models for Extremely Large-Scale Massive MIMO Channel Estimation 提出LLM4XCE，利用大语言模型解决超大规模MIMO信道估计难题 large language model
5 Experts are all you need: A Composable Framework for Large Language Model Inference 提出Comp-LLM，一种可组合的LLM推理框架，提升准确率并降低延迟。 large language model
6 EnECG: Efficient Ensemble Learning for Electrocardiogram Multi-task Foundation Model 提出EnECG，利用高效集成学习实现心电图多任务基础模型，降低计算成本。 foundation model ✅
7 Constructing Efficient Fact-Storing MLPs for Transformers 提出高效存储事实的MLP构建框架，提升Transformer的事实记忆能力 large language model
8 Orion-Bix: Bi-Axial Attention for Tabular In-Context Learning Orion-Bix：面向表格数据的双轴注意力上下文学习框架 foundation model ✅
9 CRAwDAD: Causal Reasoning Augmentation with Dual-Agent Debate 提出CRAwDAD：利用双智能体辩论增强因果推理能力 large language model

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

#题目一句话要点标签🔗⭐
10 Bridging Modalities via Progressive Re-alignment for Multimodal Test-Time Adaptation 提出BriMPR框架，通过渐进式重对齐解决多模态测试时自适应问题 contrastive learning multimodal ✅
11 PerfMamba: Performance Analysis and Pruning of Selective State Space Models PerfMamba：通过性能分析和剪枝优化选择性状态空间模型 Mamba SSM state space model
12 A Hierarchical Hybrid AI Approach: Integrating Deep Reinforcement Learning and Scripted Agents in Combat Simulations 提出分层混合AI方法，融合深度强化学习与脚本智能体，提升作战模拟性能。 reinforcement learning deep reinforcement learning
13 LFM2 Technical Report LFM2：面向边缘设备高效部署的Liquid Foundation Models，兼顾速度与性能。 curriculum learning distillation foundation model
14 SmallWorlds: Assessing Dynamics Understanding of World Models in Isolated Environments SmallWorlds：在隔离环境中评估世界模型的动态理解能力 world model state space model representation learning
15 OBLR-PO: A Theoretical Framework for Stable Reinforcement Learning 提出OBLR-PO算法，通过理论指导的自适应学习率和基线优化，提升LLM的RL后训练稳定性。 reinforcement learning large language model
16 ThetaEvolve: Test-time Learning on Open Problems ThetaEvolve：面向开放问题的测试时学习框架，实现持续进化。 reinforcement learning reward shaping large language model ✅
17 ASTRO: Adaptive Stitching via Dynamics-Guided Trajectory Rollouts ASTRO：通过动态引导的轨迹展开实现自适应拼接，提升离线强化学习性能 reinforcement learning policy learning offline RL
18 Emergent Coordination and Phase Structure in Independent Multi-Agent Reinforcement Learning 揭示独立多智能体强化学习中的涌现协调与相结构，关注规模、密度与核漂移的相互作用。 reinforcement learning IQL

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
19 Learning When to Ask: Simulation-Trained Humanoids for Mental-Health Diagnosis 提出基于模拟训练的人形机器人心理健康诊断方法，提升对话式诊断效率与安全性。 humanoid humanoid robot policy learning

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
20 Opening the Black Box: An Explainable, Few-shot AI4E Framework Informed by Physics and Expert Knowledge for Materials Engineering 提出基于物理和专家知识的AI4E框架，解决材料工程中少样本、可解释性难题 physically plausible

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

#题目一句话要点标签🔗⭐
21 Self-Supervised Dynamical System Representations for Physiological Time-Series 提出PULSE，利用自监督动态系统表征学习生理时间序列，提升泛化性和标签效率。 PULSE

⬅️ 返回 cs.LG 首页 · 🏠 返回主页