cs.LG（2024-12-06）
📊 共 13 篇论文
🎯 兴趣领域导航
支柱九：具身大模型 (Embodied Foundation Models) (7)
支柱二：RL算法与架构 (RL & Architecture) (4)
支柱六：视频提取与匹配 (Video Extraction) (2)
🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 1 | Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization | 提出WaveToken，一种基于小波变换的时间序列预测基础模型，实现高效token化。 | foundation model | ||
| 2 | The BrowserGym Ecosystem for Web Agent Research | BrowserGym生态系统：用于Web Agent研究的统一评估与开发平台 | large language model | ||
| 3 | HiVeGen -- Hierarchical LLM-based Verilog Generation for Scalable Chip Design | HiVeGen：一种基于分层LLM的可扩展芯片设计Verilog生成框架 | large language model | ||
| 4 | APOLLO: SGD-like Memory, AdamW-level Performance | APOLLO：通过近似梯度缩放，在LLM优化中实现AdamW级性能和SGD级内存效率。 | large language model | ||
| 5 | Chemist-aligned retrosynthesis by ensembling diverse inductive bias models | RetroChimera：通过集成多样归纳偏置模型实现化学家对齐的逆合成 | zero-shot transfer | ||
| 6 | Direct Quantized Training of Language Models with Stochastic Rounding | 提出基于随机舍入的语言模型直接量化训练方法，降低训练时的内存占用。 | large language model | ||
| 7 | IterL2Norm: Fast Iterative L2-Normalization | 提出IterL2Norm以解决Transformer模型的层归一化效率问题 | large language model |
🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 8 | Enhancing LLMs for Physics Problem-Solving using Reinforcement Learning with Human-AI Feedback | 提出基于人类-AI反馈强化学习的LLM物理问题求解增强方法 | reinforcement learning PPO DPO | ||
| 9 | Rethinking Time Series Forecasting with LLMs via Nearest Neighbor Contrastive Learning | 提出NNCL-TLLM，通过近邻对比学习增强LLM在时间序列预测中的性能。 | contrastive learning large language model | ||
| 10 | A Temporally Correlated Latent Exploration for Reinforcement Learning | 提出时序相关潜在探索(TeCLE)，增强强化学习在噪声环境下的探索能力 | reinforcement learning deep reinforcement learning | ||
| 11 | Putting the Iterative Training of Decision Trees to the Test on a Real-World Robotic Task | 提出迭代训练决策树算法，成功应用于真实机器人倒立摆控制任务。 | reinforcement learning deep reinforcement learning DRL |
🔬 支柱六：视频提取与匹配 (Video Extraction) (2 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 12 | Multi-Objective Alignment of Large Language Models Through Hypervolume Maximization | 提出HaM算法，通过最大化超体积实现大语言模型的多目标对齐 | HuMoR large language model | ||
| 13 | Learning Hidden Physics and System Parameters with Deep Operator Networks | 提出深度算子网络以解决隐藏物理和系统参数识别问题 | sparse sensors |