cs.LG（2025-04-19）

📊 共 11 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) (6 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (4) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
1 DConAD: A Differencing-based Contrastive Representation Learning Framework for Time Series Anomaly Detection DConAD：基于差分对比表示学习的时间序列异常检测框架 representation learning contrastive learning spatiotemporal ✅
2 Accelerating LLM Inference with Flexible N:M Sparsity via A Fully Digital Compute-in-Memory Accelerator 提出FLOW与FlexCiM，加速具有灵活N:M稀疏性的大语言模型推理。 SSM large language model foundation model ✅
3 SRPO: A Cross-Domain Implementation of Large-Scale Reinforcement Learning on LLM 提出SRPO，通过两阶段训练和历史重采样提升LLM在数学推理和代码生成上的能力 reinforcement learning large language model
4 Optimal Lattice Boltzmann Closures through Multi-Agent Reinforcement Learning 提出多智能体强化学习以优化格子玻尔兹曼闭合模型 reinforcement learning spatiotemporal
5 Improving RL Exploration for LLM Reasoning through Retrospective Replay 提出基于回溯重放的强化学习(RRL)，提升LLM在复杂推理任务中的探索能力。 reinforcement learning RLHF large language model
6 Learning from Stochastic Teacher Representations Using Student-Guided Knowledge Distillation 提出基于学生引导知识蒸馏的随机自蒸馏方法，提升资源受限场景下的模型性能。 distillation

🔬 支柱九：具身大模型 (Embodied Foundation Models) (4 篇)

#题目一句话要点标签🔗⭐
7 Integrating Single-Cell Foundation Models with Graph Neural Networks for Drug Response Prediction 利用单细胞Foundation模型和图神经网络进行药物反应预测 foundation model
8 A Physics-guided Multimodal Transformer Path to Weather and Climate Sciences 提出基于物理引导的多模态Transformer框架，用于提升天气和气候科学模型的性能。 multimodal
9 Do You Really Need Public Data? Surrogate Public Data for Differential Privacy on Tabular Data 提出利用模式信息的代理公共数据，解决表格数据差分隐私学习中公共数据依赖问题 large language model
10 FGMP: Fine-Grained Mixed-Precision Weight and Activation Quantization for Hardware-Accelerated LLM Inference 提出FGMP以解决大语言模型推理效率问题 large language model

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
11 Learning Enhanced Structural Representations with Block-Based Uncertainties for Ocean Floor Mapping 提出基于分块不确定性的结构化表征学习方法，用于提升海底地形测绘精度。 VQ-VAE

⬅️ 返回 cs.LG 首页 · 🏠 返回主页