cs.LG（2025-04-29）

📊 共 17 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) (9) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗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 GLIP-OOD: Zero-Shot Graph OOD Detection with Graph Foundation Model 提出GLIP-OOD，利用图基础模型实现零样本图OOD检测，无需ID数据训练。 large language model foundation model
2 Graph Synthetic Out-of-Distribution Exposure with Large Language Models 提出GOE-LLM框架，利用大语言模型进行图结构OOD检测，无需真实OOD数据。 large language model
3 A Survey on Parameter-Efficient Fine-Tuning for Foundation Models in Federated Learning 联邦学习中面向大模型的参数高效微调方法综述 foundation model
4 NeuRel-Attack: Neuron Relearning for Safety Disalignment in Large Language Models NeuRel-Attack：通过神经元重学习实现大语言模型安全性解除 large language model
5 A Cost-Effective LLM-based Approach to Identify Wildlife Trafficking in Online Marketplaces 提出一种低成本的基于LLM的方法，用于识别在线市场中的野生动物非法交易。 large language model
6 LIFT: LLM-Based Pragma Insertion for HLS via GNN Supervised Fine-Tuning LIFT：基于LLM和GNN监督微调的HLS编译pragma自动插入方法 large language model
7 ACE: A Security Architecture for LLM-Integrated App Systems ACE：为LLM集成应用系统提供安全保障的架构 large language model
8 Combatting Dimensional Collapse in LLM Pre-Training Data via Diversified File Selection 提出DiSF算法，通过多样化文件选择解决LLM预训练数据中的维度坍塌问题。 large language model
9 GaLore 2: Large-Scale LLM Pre-Training by Gradient Low-Rank Projection GaLore 2：通过梯度低秩投影实现大规模LLM预训练，解决内存瓶颈。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
10 Reinforcement Learning for Reasoning in Large Language Models with One Training Example 提出单样本强化学习与可验证奖励(1-shot RLVR)，提升大语言模型数学推理能力。 reinforcement learning PPO large language model ✅
11 Toward Efficient Exploration by Large Language Model Agents 提出基于LLM的后验采样强化学习方法，提升自然语言任务中的探索效率 reinforcement learning large language model
12 Token-Efficient RL for LLM Reasoning 提出Token高效强化学习方法，解决LLM推理中内存和计算资源限制问题 reinforcement learning large language model
13 Q-Fusion: Diffusing Quantum Circuits Q-Fusion：提出基于扩散模型的量子电路生成方法，解决量子架构搜索难题。 reinforcement learning large language model
14 Quantum-Enhanced Hybrid Reinforcement Learning Framework for Dynamic Path Planning in Autonomous Systems 提出量子增强混合强化学习框架，用于自主系统动态路径规划 reinforcement learning
15 Representation Learning Preserving Ignorability and Covariate Matching for Treatment Effects 提出一种新的表征学习方法，同时解决因果效应估计中的混淆偏差和协变量失配问题。 representation learning
16 Group Relative Knowledge Distillation: Learning from Teacher's Relational Inductive Bias 提出组相对知识蒸馏(GRKD)，利用教师模型的相对关系归纳偏置提升学生模型泛化能力。 distillation

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

#题目一句话要点标签🔗⭐
17 Efficient LLMs with AMP: Attention Heads and MLP Pruning 提出AMP：一种高效的LLM剪枝方法，用于加速推理并降低资源消耗 AMP large language model

⬅️ 返回 cs.LG 首页 · 🏠 返回主页