cs.LG（2024-04-12）

📊 共 21 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (12 🔗3) 支柱九：具身大模型 (Embodied Foundation Models) (8) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (12 篇)

#题目一句话要点标签🔗⭐
1 RLHF Deciphered: A Critical Analysis of Reinforcement Learning from Human Feedback for LLMs 深入分析RLHF以优化大语言模型的训练方法 reinforcement learning RLHF large language model
2 Advancing Forest Fire Prevention: Deep Reinforcement Learning for Effective Firebreak Placement 提出深度强化学习方法以优化防火隔离带布局 reinforcement learning deep reinforcement learning
3 Dataset Reset Policy Optimization for RLHF 提出数据集重置策略优化以提升RLHF性能 reinforcement learning PPO RLHF ✅
4 Kinematics Modeling of Peroxy Free Radicals: A Deep Reinforcement Learning Approach 利用深度强化学习预测过氧自由基的反应动力学 reinforcement learning deep reinforcement learning
5 The Illusion of State in State-Space Models 揭示状态空间模型在状态跟踪中的局限性 Mamba SSM large language model
6 Adversarial Imitation Learning via Boosting 提出AILBoost以解决现有对抗模仿学习的不足问题 policy learning imitation learning
7 TSLANet: Rethinking Transformers for Time Series Representation Learning 提出TSLANet以解决时间序列数据表示学习中的噪声敏感性问题 representation learning ✅
8 Anti-Byzantine Attacks Enabled Vehicle Selection for Asynchronous Federated Learning in Vehicular Edge Computing 提出基于深度强化学习的车辆选择方案以应对异步联邦学习中的拜占庭攻击问题 reinforcement learning deep reinforcement learning DRL
9 SIR-RL: Reinforcement Learning for Optimized Policy Control during Epidemiological Outbreaks in Emerging Market and Developing Economies 提出SIR-RL框架以优化疫情期间公共卫生与经济稳定的政策控制 reinforcement learning
10 Exploring Contrastive Learning for Long-Tailed Multi-Label Text Classification 提出对比学习方法以解决长尾多标签文本分类问题 contrastive learning
11 Generalized Population-Based Training for Hyperparameter Optimization in Reinforcement Learning 提出广义种群训练以优化强化学习中的超参数 reinforcement learning
12 Megalodon: Efficient LLM Pretraining and Inference with Unlimited Context Length 提出Megalodon以解决长序列建模效率问题 state space model linear attention ✅

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

#题目一句话要点标签🔗⭐
13 Experimental Design for Active Transductive Inference in Large Language Models 提出主动上下文提示设计以优化大语言模型推理性能 large language model
14 CATS: Contextually-Aware Thresholding for Sparsity in Large Language Models 提出CATS框架以降低大语言模型的推理成本 large language model
15 Differentiable and Stable Long-Range Tracking of Multiple Posterior Modes 提出一种可微分且稳定的多后验模式长程跟踪方法 multimodal
16 Training a Vision Language Model as Smartphone Assistant 提出视觉语言模型以解决智能手机助手任务执行问题 large language model
17 AdapterSwap: Continuous Training of LLMs with Data Removal and Access-Control Guarantees 提出AdapterSwap以解决LLMs数据动态管理问题 large language model
18 Reducing hallucination in structured outputs via Retrieval-Augmented Generation 通过检索增强生成技术减少结构化输出中的幻觉问题 large language model
19 FLoRA: Enhancing Vision-Language Models with Parameter-Efficient Federated Learning 提出FLoRA以解决视觉语言模型的隐私与效率问题 multimodal
20 Conformal Prediction via Regression-as-Classification 通过回归转分类提出保形预测方法以解决回归问题 multimodal

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
21 Hindsight PRIORs for Reward Learning from Human Preferences 提出Hindsight PRIOR以解决奖励学习中的信用分配问题 locomotion manipulation reinforcement learning ✅

⬅️ 返回 cs.LG 首页 · 🏠 返回主页