LLM · Agent | 使用 LLM agent 玩各种游戏
这个 repo 总结了 LLM agents play games 的论文,最近读了一些。
论文列表:
- Plan4MC: Skill Reinforcement Learning and Planning for Open-World Minecraft Tasks
- 最初发表时间:2023.03
- arxiv:https://arxiv.org/abs/2303.16563
- GitHub:https://github.com/PKU-RL/Plan4MC
- 网站:https://sites.google.com/view/plan4mc
- 投了 NeurIPS 2023 Workshop FMDM,ICLR 2024 拒稿。
- 本站博客:LLM · RL | Plan4MC:使用有向无环图 high-level planning + 基于 RL 执行 low-level policy
- Building Cooperative Embodied Agents Modularly with Large Language Models (ICLR 2024)
- 最初发表时间:2023.07
- arxiv:https://arxiv.org/abs/2307.02485
- GitHub:https://github.com/UMass-Embodied-AGI/Co-LLM-Agents
- ICLR 2024 poster。
- 本站博客:LLM · Agent | 记忆模块 + 交流模块,让 agent 合作完成复杂任务
- Avalon's Game of Thoughts: Battle Against Deception through Recursive Contemplation
- 最初发表时间:2023.10
- arxiv:https://arxiv.org/abs/2310.01320
- GitHub:https://github.com/Shenzhi-Wang/recon
- 网站:https://shenzhi-wang.github.io/avalon_recon/
- ICLR 2024 withdraw。
- 本站博客:LLM · Agent | 通过推断别人身份 + 别人对自己说话的看法,让 agent 在阿瓦隆中欺骗
- Language Agents with Reinforcement Learning for Strategic Play in the Werewolf Game
- 最初发表时间:2023.10
- arxiv:https://arxiv.org/abs/2310.18940
- ICLR 2024 拒稿。
- 本站博客:
- Large Language Models Play StarCraft II: Benchmarks and A Chain of Summarization Approach
- 最初发表时间:2023.12
- arxiv:https://arxiv.org/abs/2312.11865
- GitHub:https://github.com/histmeisah/Large-Language-Models-play-StarCraftII/
- 网站:https://sites.google.com/view/textstarcraft2
- NeurIPS 2024 poster。
- 本站博客:LLM · Agent | 使用 LLM 的通识决策能力,玩星际争霸
- PokerGPT: An End-to-End Lightweight Solver for Multi-Player Texas Hold'em via Large Language Model
- 最初发表时间:2024.01
- arxiv:https://arxiv.org/abs/2401.06781
- 开源代码 404 了。好像没有投任何会议。
- 本站博客:
- SwarmBrain: Embodied agent for real-time strategy game StarCraft II via large language models
- 最初发表时间:2024.01
- arxiv:https://arxiv.org/abs/2401.17749
- 宝马公司(BMW)写的神秘文章,好像没有投任何会议。
- 本站博客:
- Odyssey: Empowering Minecraft Agents with Open-World Skills
- 最初发表时间:2024.07
- arxiv:https://arxiv.org/abs/2407.15325
- GitHub:https://github.com/zju-vipa/Odyssey
- ICLR 2025 withdraw。
- 本站博客:
- Agent Planning with World Knowledge Model (NeurIPS 2024)
- 最初发表时间:2024.10
- arxiv:https://arxiv.org/abs/2405.14205
- GitHub:https://github.com/zjunlp/WKM
- hugging face:https://huggingface.co/collections/zjunlp/wkm-6684c611102213b6d8104f84
- NeurIPS 2024 poster。
- 本站博客:
- https://zhuanlan.zhihu.com/p/29450658588