cs.LG（2024-09-20）

📊 共 16 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (8) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (8 篇)

#题目一句话要点标签🔗⭐
1 ChemDFM-X: Towards Large Multimodal Model for Chemistry 提出ChemDFM-X化学交叉模态对话大模型，弥合化学数据模态鸿沟。 foundation model multimodal
2 Continual Learning for Multimodal Data Fusion of a Soft Gripper 提出一种基于原型存储的增量式多模态融合学习算法，用于软体抓取器。 multimodal
3 Prithvi WxC: Foundation Model for Weather and Climate Prithvi WxC：用于天气和气候的23亿参数开源基础模型，支持多种下游任务。 foundation model
4 Towards Long-Context Time Series Foundation Models 提出一种压缩记忆机制，扩展时间序列基础模型处理长上下文多元时间序列的能力。 foundation model
5 OATS: Outlier-Aware Pruning Through Sparse and Low Rank Decomposition OATS：通过稀疏低秩分解实现异常值感知的模型剪枝 large language model foundation model
6 BoilerTAI: A Platform for Enhancing Instruction Using Generative AI in Educational Forums BoilerTAI：一个利用生成式AI增强教育论坛教学的平台 large language model
7 ControlMath: Controllable Data Generation Promotes Math Generalist Models 提出ControlMath，通过可控数据生成提升数学通用模型能力 large language model
8 An Adaptive End-to-End IoT Security Framework Using Explainable AI and LLMs 提出基于XAI和LLM的自适应端到端物联网安全框架，提升威胁检测与响应能力。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#题目一句话要点标签🔗⭐
9 Optimizing RLHF Training for Large Language Models with Stage Fusion RLHFuse：通过阶段融合优化大型语言模型的RLHF训练 reinforcement learning RLHF large language model
10 State space models, emergence, and ergodicity: How many parameters are needed for stable predictions? 提出参数阈值理论以解决线性动态系统学习问题 state space model large language model
11 Higher-Order Message Passing for Glycan Representation Learning 提出基于高阶消息传递的组合复形网络，用于糖链表示学习，刷新GlycanML benchmark性能。 predictive model representation learning
12 Federated Learning with Label-Masking Distillation 提出FedLMD，通过标签掩码蒸馏解决联邦学习中的标签分布倾斜问题 distillation ✅
13 A Generative Framework for Predictive Modeling of Multiple Chronic Conditions Using Graph Variational Autoencoder and Bandit-Optimized Graph Neural Network 提出生成框架以解决多重慢性病预测问题 predictive model

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
14 Inductive Spatial Temporal Prediction Under Data Drift with Informative Graph Neural Network 提出INF-GNN，解决数据漂移下归纳时空预测的泛化性问题。 spatial relationship TAMP

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
15 Learning to Play Video Games with Intuitive Physics Priors 提出基于直观物理先验的视频游戏学习方法，提升泛化性。 affordance

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

#题目一句话要点标签🔗⭐
16 Efficient Training of Deep Neural Operator Networks via Randomized Sampling 提出基于随机采样的DeepONet训练方法，提升泛化能力并加速复杂动力学预测。 spatiotemporal

⬅️ 返回 cs.LG 首页 · 🏠 返回主页