Elamur Research⌕
☰
home / papers / 2023 / GameGPT: Multi-agent Collaborative Framework for Game Development

GameGPT: Multi-agent Collaborative Framework for Game Development

Paper · 2023 · Dake Chen, Haoyang Zhang, Hanbin Wang et al.

Summary

GameGPT is a framework that automates game development with LLM-based agents. Besides hallucination, the authors identify a second weakness they call redundancy: models tend to produce superfluous, uninformative tasks or code snippets. The system divides work among specialized agents. Each agent has its own private memory and also reads a shared public record. Development runs through five stages: planning, task classification, code generation, execution, and result summarization.

The mitigations are layered. "Dual collaboration" pairs LLMs with small expert models and pairs executing agents with reviewer agents. Genre-specific plan templates and instruction tuning on in-house lexicons constrain what the models produce. Long Lua scripts are split into short snippets, which allows prompting with example snippets and choosing among several generated candidates. Users can inspect and correct intermediate outputs through a frontend.

The authors claim better decision-making and error correction, but the provided text contains no quantitative evaluation. The work is mainly useful as a design blueprint for role-based multi-agent pipelines in domain-specific code generation.

Key points

  • The framework uses eight roles: a designer responsible for game content, a manager who oversees game development, a reviewer who checks plans, an engineer who develops the game, a reviewer of tasks, an engineer for the game engine, a reviewer of code, and an engineer who tests the game engine. Reviewer/critic agents check the first three stages.
  • Planning combines four orthogonal measures. Requests are classified into one of five supported genres and filled into a genre template. A plan reviewer agent checks the plan, the manager LLM is instruction-tuned on in-house request–plan pairs written by game practitioners, and users can edit the plan interactively.
  • In task classification, two expert models determine each task's type and arguments. The task reviewer then builds a directed acyclic graph of task dependencies and traverses it with breadth-first search to fix the execution order. When arguments cannot be inferred, it asks the user for them.
  • Code generation decouples Lua scripts into short snippets and adds in-context examples at the snippet level. It generates K candidates per task, which are tested in a virtual environment and filtered with user feedback. The model is also fine-tuned on an in-house lexicon of annotated game code snippets.
  • A testing engineer agent sends each task's code to the game engine in the planned order. It collects logs and tracebacks and presents a final summary to the user. No benchmark numbers appear in the available text.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Tsinghua UniversityBeijing · China439 works · 94,783 citations🎓University of Southern CaliforniaLos Angeles · United States183 works · 37,526 citations

 Fields, topics and concepts

📚 Shelves 3The field of science this work belongs to (OpenAlex fields).Computer Science · 19,503 worksArtificial Intelligence · 18,805 worksInformation Systems · 1,888 works

Cite

Cite as Dake Chen, Haoyang Zhang, Hanbin Wang, Yunhao Huo, Yuzhao Li and Junjie Wang. 2023. GameGPT: Multi-agent Collaborative Framework for Game Development. arXiv:2310.08067.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2310.08067 · arXiv 2310.08067

@misc{chen2023gamegpt,
  title = {{GameGPT: Multi-agent Collaborative Framework for Game Development}},
  author = {Dake Chen, Haoyang Zhang, Hanbin Wang, Yunhao Huo, Yuzhao Li and Junjie Wang},
  year = {2023},
  doi = {10.48550/arxiv.2310.08067},
  eprint = {2310.08067},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2310.08067}
}

References

24 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. J. Laird and M. VanLent, “Human-Level AI’s Killer Application: Interactive Computer Games,” AI magazine, vol. 22, no. 2, pp. 15–15, 2001. find
  2. D. Perez-Liebana, S. Samothrakis, J. Togelius, T. Schaul, and S. Lucas, “General Video Game Ai: Competition, Challenges And Opportunities,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 30, no. 1, 2016. find
  3. J. H. Kim and R. Wu, “Leveraging machine learning for game development,” 2021. [Online]. Available: https://blog.research.google/2021/03/leveraging-machine-learning-for-game.html?m=1 ↗
  4. C. Berner, G. Brockman, B. Chan, V. Cheung, P. D˛ebiak, C. Dennison, D. Farhi, Q. Fischer, S. Hashme, C. Hesse et al., “Dota 2 With Large Scale Deep Reinforcement Learning,” arXiv preprint arXiv:1912.06680, 2019. find
  5. S. Sudhakaran, M. González-Duque, C. Glanois, M. Freiberger, E. Najarro, and S. Risi, “MarioGPT: Open-Ended Text2Level Generation Through Large Language Models,” 2023. find
  6. A. Radford, J. Wu, R. Child, D. Luan, D. Amodei, I. Sutskever et al., “Language Models Are Unsupervised Multitask Learners,” OpenAI blog, vol. 1, no. 8, p. 9, 2019. find
  7. A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention Is All You Need,” Advances in neural information processing systems, vol. 30, 2017.
  8. A. Radford, K. Narasimhan, T. Salimans, I. Sutskever et al., “Improving Language Understanding By Generative Pre-Training,” 2018. find
  9. J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-Training Of Deep Bidirectional Transformers For Language Understanding,” arXiv preprint arXiv:1810.04805, 2018.
  10. T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell et al., “Language Models Are Few-Shot Learners,” Advances in neural information processing systems, vol. 33, pp. 1877–1901, 2020.
  11. H. Touvron, T. Lavril, G. Izacard, X. Martinet, M.-A. Lachaux, T. Lacroix, B. Rozière, N. Goyal, E. Hambro, F. Azhar et al., “Llama: Open And Efficient Foundation Language Models,” arXiv preprint arXiv:2302.13971, 2023.
  12. W.-L. Chiang, Z. Li, Z. Lin, Y. Sheng, Z. Wu, H. Zhang, L. Zheng, S. Zhuang, Y. Zhuang, J. E. Gonzalez et al., “Vicuna: An Open-Source Chatbot Impressing Gpt-4 With 90%* Chatgpt Quality,” See https://vicuna.lmsys. org (accessed 14 April 2023), 2023. ↗
  13. D. Hendrycks, C. Burns, S. Basart, A. Zou, M. Mazeika, D. Song, and J. Steinhardt, “Measuring Massive Multitask Language Understanding,” arXiv preprint arXiv:2009.03300, 2020.
  14. H. W. Chung, L. Hou, S. Longpre, B. Zoph, Y. Tay, W. Fedus, E. Li, X. Wang, M. Dehghani, S. Brahma et al., “Scaling Instruction-Finetuned Language Models,” arXiv preprint arXiv:2210.11416, 2022. find
  15. Y. Wang, Y. Kordi, S. Mishra, A. Liu, N. A. Smith, D. Khashabi, and H. Hajishirzi, “Self-Instruct: Aligning Language Model With Self Generated Instructions,” arXiv preprint arXiv:2212.10560, 2022. find
  16. C. Xu, Q. Sun, K. Zheng, X. Geng, P. Zhao, J. Feng, C. Tao, and D. Jiang, “Wizardlm: Empowering Large Language Models To Follow Complex Instructions,” arXiv preprint arXiv:2304.12244, 2023.
  17. L. Ouyang, J. Wu, X. Jiang, D. Almeida, C. Wainwright, P. Mishkin, C. Zhang, S. Agarwal, K. Slama, A. Ray et al., “Training Language Models To Follow Instructions With Human Feedback,” Advances in Neural Information Processing Systems, vol. 35, pp. 27 730–27 744, 2022.
  18. H. Yang, S. Yue, and Y. He, “Auto-GPT For Online Decision Making: Benchmarks And Additional Opinions,” arXiv preprint arXiv:2306.02224, 2023.
  19. Y. Shen, K. Song, X. Tan, D. Li, W. Lu, and Y. Zhuang, “Hugginggpt: Solving Ai Tasks With Chatgpt And Its Friends In Huggingface,” arXiv preprint arXiv:2303.17580, 2023.
  20. C. Qian, X. Cong, C. Yang, W. Chen, Y. Su, J. Xu, Z. Liu, and M. Sun, “Communicative Agents For Software Development,” arXiv preprint arXiv:2307.07924, 2023.
  21. S. Hong, X. Zheng, J. Chen, Y. Cheng, C. Zhang, Z. Wang, S. K. S. Yau, Z. Lin, L. Zhou, C. Ran et al., “MetaGPT: Meta Programming For Multi-Agent Collaborative Framework,” arXiv preprint arXiv:2308.00352, 2023.
  22. J. S. Park, J. C. O’Brien, C. J. Cai, M. R. Morris, P. Liang, and M. S. Bernstein, “Generative Agents: Interactive Simulacra Of Human Behavior,” arXiv preprint arXiv:2304.03442, 2023. find
  23. G. Li, H. A. A. K. Hammoud, H. Itani, D. Khizbullin, and B. Ghanem, “CAMEL: Communicative Agents For "Mind" Exploration Of Large Scale Language Model Society,” 2023. find
  24. R. Gong, Q. Huang, X. Ma, H. Vo, Z. Durante, Y. Noda, Z. Zheng, S.-C. Zhu, D. Terzopoulos, L. Fei-Fei et al., “MindAgent: Emergent Gaming Interaction,” arXiv preprint arXiv:2309.09971, 2023. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/chen-2023-gamegpt-multi-agent-collaborative-framework/