cs.LG（2024-08-20）

📊 共 16 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) (5 🔗1) 支柱五：交互与反应 (Interaction & Reaction) (1 🔗1) 支柱四：生成式动作 (Generative Motion) (1)

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

#题目一句话要点标签🔗⭐
1 Learning Multimodal Latent Space with EBM Prior and MCMC Inference 提出EBM先验与MCMC推理的多模态隐空间学习方法，提升跨模态生成效果 multimodal
2 Towards Foundation Models for the Industrial Forecasting of Chemical Kinetics 提出基于MLP-Mixer的工业化学动力学预测基础模型方法 foundation model
3 AnyGraph: Graph Foundation Model in the Wild AnyGraph：面向通用图学习的图基础模型，解决异构图数据泛化难题。 foundation model
4 CoRA: Collaborative Information Perception by Large Language Model's Weights for Recommendation 提出CoRA以解决大语言模型推荐中的协同信息整合问题 large language model
5 LLM-Barber: Block-Aware Rebuilder for Sparsity Mask in One-Shot for Large Language Models LLM-Barber：一种面向大语言模型的一次性块感知稀疏掩码重建方法 large language model ✅
6 Do Neural Scaling Laws Exist on Graph Self-Supervised Learning? 揭示图自监督学习的规模定律缺失：现有方法难以支撑图基础模型的构建 foundation model ✅
7 A Little Confidence Goes a Long Way 提出基于LLM隐层激活探针的二分类方法，在低计算资源下实现媲美大型LLM的性能。 large language model
8 DOMBA: Double Model Balancing for Access-Controlled Language Models via Minimum-Bounded Aggregation 提出DOMBA：通过最小有界聚合的双模型平衡方法，用于访问控制语言模型。 large language model
9 Tracing Privacy Leakage of Language Models to Training Data via Adjusted Influence Functions 提出启发式调整的影响函数(HAIF)，以更精确地追踪语言模型中的隐私泄露。 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🔗⭐
10 Integrating Multi-Modal Input Token Mixer Into Mamba-Based Decision Models: Decision MetaMamba 提出Decision MetaMamba，通过多模态输入Token Mixer提升Mamba在离线强化学习中的决策能力。 reinforcement learning offline RL offline reinforcement learning
11 An End-to-End Reinforcement Learning Based Approach for Micro-View Order-Dispatching in Ride-Hailing 提出基于端到端强化学习的D2SN模型，解决网约车微观视角下的订单分配问题。 reinforcement learning spatial relationship spatiotemporal
12 Offline Model-Based Reinforcement Learning with Anti-Exploration 提出MoMo：一种基于反探索的离线模型强化学习算法，提升D4RL数据集性能。 reinforcement learning offline RL offline reinforcement learning
13 Centralized Reward Agent for Knowledge Sharing and Transfer in Multi-Task Reinforcement Learning 提出集中式奖励代理CRA，用于多任务强化学习中的知识共享与迁移。 reinforcement learning reward shaping
14 Universal Novelty Detection Through Adaptive Contrastive Learning 提出基于自适应对比学习的通用新颖性检测方法UNODE，提升模型在不同分布数据上的泛化能力。 contrastive learning ✅

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

#题目一句话要点标签🔗⭐
15 SubgoalXL: Subgoal-based Expert Learning for Theorem Proving SubgoalXL：基于子目标的专家学习提升LLM定理证明能力 IMoS large language model ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
16 Atmospheric Transport Modeling of CO$_2$ with Neural Networks 利用深度神经网络进行大气CO₂输送建模，实现长期稳定和质量守恒的模拟。 physically plausible

⬅️ 返回 cs.LG 首页 · 🏠 返回主页