Learning Game-Playing Agents with Generative Code Optimization
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912556123160576 |
|---|---|
| author | Kuang, Zhiyi Rong, Ryan Yuan, YuCheng Nie, Allen |
| author_facet | Kuang, Zhiyi Rong, Ryan Yuan, YuCheng Nie, Allen |
| contents | We present a generative optimization approach for learning game-playing agents, where policies are represented as Python programs and refined using large language models (LLMs). Our method treats decision-making policies as self-evolving code, with current observation as input and an in-game action as output, enabling agents to self-improve through execution traces and natural language feedback with minimal human intervention. Applied to Atari games, our game-playing Python program achieves performance competitive with deep reinforcement learning (RL) baselines while using significantly less training time and much fewer environment interactions. This work highlights the promise of programmatic policy representations for building efficient, adaptable agents capable of complex, long-horizon reasoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_19506 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning Game-Playing Agents with Generative Code Optimization Kuang, Zhiyi Rong, Ryan Yuan, YuCheng Nie, Allen Machine Learning Artificial Intelligence We present a generative optimization approach for learning game-playing agents, where policies are represented as Python programs and refined using large language models (LLMs). Our method treats decision-making policies as self-evolving code, with current observation as input and an in-game action as output, enabling agents to self-improve through execution traces and natural language feedback with minimal human intervention. Applied to Atari games, our game-playing Python program achieves performance competitive with deep reinforcement learning (RL) baselines while using significantly less training time and much fewer environment interactions. This work highlights the promise of programmatic policy representations for building efficient, adaptable agents capable of complex, long-horizon reasoning. |
| title | Learning Game-Playing Agents with Generative Code Optimization |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2508.19506 |