cs.LG（2024-08-19）

📊 共 18 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) 支柱九：具身大模型 (Embodied Foundation Models) (8 🔗2) 支柱四：生成式动作 (Generative Motion) (1)

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

#题目一句话要点标签🔗⭐
1 Personalizing Reinforcement Learning from Human Feedback with Variational Preference Learning 提出基于变分偏好学习的个性化RLHF方法，解决用户偏好多样性问题 reinforcement learning preference learning RLHF
2 Molecular Graph Representation Learning Integrating Large Language Models with Domain-specific Small Models 提出MolGraph-LarDo框架，融合大语言模型与领域小模型提升分子图表示学习 representation learning large language model
3 Efficient Exploration in Deep Reinforcement Learning: A Novel Bayesian Actor-Critic Algorithm 提出一种新型贝叶斯Actor-Critic算法，提升深度强化学习中的高效探索能力。 reinforcement learning deep reinforcement learning DRL
4 Transformers to SSMs: Distilling Quadratic Knowledge to Subquadratic Models 提出MOHAWK方法以将Transformer知识蒸馏至子二次模型 Mamba SSM state space model
5 Leveraging Superfluous Information in Contrastive Representation Learning 提出SuperInfo损失函数，通过区分预测性和冗余信息提升对比学习表征 representation learning contrastive learning
6 Data Augmentation of Contrastive Learning is Estimating Positive-incentive Noise 提出基于π-噪声生成器的对比学习数据增强框架，提升模型性能 contrastive learning
7 Liquid Fourier Latent Dynamics Networks for fast GPU-based numerical simulations in computational cardiology 提出Liquid Fourier LDNets，加速计算心脏病学中基于GPU的数值模拟。 latent dynamics
8 Structure-enhanced Contrastive Learning for Graph Clustering 提出结构增强对比学习(SECL)用于提升图聚类性能 contrastive learning
9 GARLIC: GPT-Augmented Reinforcement Learning with Intelligent Control for Vehicle Dispatching GARLIC：基于GPT增强强化学习的智能车辆调度框架 reinforcement learning

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

#题目一句话要点标签🔗⭐
10 Enhance Modality Robustness in Text-Centric Multimodal Alignment with Adversarial Prompting 提出基于对抗提示的文本中心多模态对齐方法，增强模态鲁棒性 large language model foundation model multimodal
11 SMILE: Zero-Shot Sparse Mixture of Low-Rank Experts Construction From Pre-Trained Foundation Models SMILE：基于预训练模型零样本构建稀疏低秩专家混合模型 large language model foundation model ✅
12 MoDeGPT: Modular Decomposition for Large Language Model Compression MoDeGPT：通过模块分解实现大语言模型高效压缩，无需微调。 large language model
13 Understanding Generative AI Content with Embedding Models 利用嵌入模型理解生成式AI内容，揭示真实样本与AI生成样本的内在可分性 foundation model
14 In-Context Learning with Representations: Contextual Generalization of Trained Transformers 研究Transformer在上下文学习中泛化能力，证明其可学习模板信息以推广到未见示例和任务。 large language model
15 Strategic Demonstration Selection for Improved Fairness in LLM In-Context Learning 提出基于聚类和演化策略的ICL提示选择方法，提升LLM在表格数据上的公平性 large language model
16 Icing on the Cake: Automatic Code Summarization at Ericsson 在爱立信公司，研究了基于LLM的Java方法自动代码摘要生成，并提出了更轻量级的方法。 large language model
17 AdapMoE: Adaptive Sensitivity-based Expert Gating and Management for Efficient MoE Inference AdapMoE：面向高效MoE推理的自适应专家门控与管理 large language model ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
18 Augmenting train maintenance technicians with automated incident diagnostic suggestions 提出一种自动事件诊断建议系统，辅助列车维护技师提升效率。 physically plausible

⬅️ 返回 cs.LG 首页 · 🏠 返回主页