cs.LG（2024-03-18）
📊 共 21 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) (7 🔗2)
支柱八：物理动画 (Physics-based Animation) (3)
支柱四：生成式动作 (Generative Motion) (2)
🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 1 | Reinforcement Learning from Delayed Observations via World Models | 提出世界模型以解决延迟观察的强化学习问题 | reinforcement learning world model world models | ||
| 2 | Supervised Fine-Tuning as Inverse Reinforcement Learning | 提出基于专家示范的数据集对齐方法以解决LLM对齐问题 | reinforcement learning imitation learning inverse reinforcement learning | ||
| 3 | Unveil Conditional Diffusion Models with Classifier-free Guidance: A Sharp Statistical Theory | 提出条件扩散模型的统计理论以解决生成样本的复杂性问题 | reinforcement learning classifier-free guidance | ||
| 4 | Offline Multitask Representation Learning for Reinforcement Learning | 提出MORL算法以解决离线多任务表示学习问题 | reinforcement learning representation learning | ||
| 5 | S-JEPA: towards seamless cross-dataset transfer through dynamic spatial attention | 提出Signal-JEPA以解决EEG信号跨数据集迁移问题 | JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture | ||
| 6 | Investigating the Benefits of Projection Head for Representation Learning | 提出投影头以提升表征学习效果 | representation learning contrastive learning | ||
| 7 | The Value of Reward Lookahead in Reinforcement Learning | 提出未来奖励预见性以提升强化学习性能 | reinforcement learning offline RL | ||
| 8 | Towards Global Optimality for Practical Average Reward Reinforcement Learning without Mixing Time Oracles | 提出多层次演员-评论家框架以解决平均奖励强化学习中的混合时间问题 | reinforcement learning | ||
| 9 | Semantic-Enhanced Representation Learning for Road Networks with Temporal Dynamics | 提出Toast和DyToast以增强道路网络的时序动态表示学习 | representation learning |
🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 10 | Compiler generated feedback for Large Language Models | 提出基于大型语言模型的编译器反馈优化LLVM汇编代码 | large language model | ||
| 11 | Accelerating Scientific Discovery with Generative Knowledge Extraction, Graph-Based Representation, and Multimodal Intelligent Graph Reasoning | 通过生成知识提取与图形推理加速科学发现 | multimodal | ||
| 12 | Larimar: Large Language Models with Episodic Memory Control | 提出Larimar以解决大语言模型知识更新效率问题 | large language model | ✅ | |
| 13 | Narrative Feature or Structured Feature? A Study of Large Language Models to Identify Cancer Patients at Risk of Heart Failure | 利用大语言模型识别癌症患者心衰风险 | large language model | ||
| 14 | RouterBench: A Benchmark for Multi-LLM Routing System | 提出RouterBench以解决多LLM路由系统评估问题 | large language model | ✅ | |
| 15 | Linguacodus: A Synergistic Framework for Transformative Code Generation in Machine Learning Pipelines | 提出Linguacodus框架以解决自然语言到代码生成的挑战 | large language model | ||
| 16 | LLM Guided Evolution -- The Automation of Models Advancing Models | 提出引导进化框架以提升模型自动化设计效率 | large language model |
🔬 支柱八：物理动画 (Physics-based Animation) (3 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 17 | Causality-Aware Spatiotemporal Graph Neural Networks for Spatiotemporal Time Series Imputation | 提出因果感知时空图神经网络以解决时空时间序列缺失值问题 | spatiotemporal | ||
| 18 | Large-scale flood modeling and forecasting with FloodCast | 提出FloodCast以解决大规模洪水建模与预测问题 | spatiotemporal | ||
| 19 | Adaptive LPD Radar Waveform Design with Generative Deep Learning | 提出基于生成深度学习的自适应低探测概率雷达波形设计 | PULSE |
🔬 支柱四：生成式动作 (Generative Motion) (2 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 20 | PITA: Physics-Informed Trajectory Autoencoder | 提出物理信息轨迹自编码器PITA以解决轨迹生成的物理合理性问题 | physically plausible | ||
| 21 | HyperVQ: MLR-based Vector Quantization in Hyperbolic Space | 提出HyperVQ以解决欧几里得空间向量量化的局限性 | VQ-VAE |