Superhuman AI for Stratego Using Self-Play Reinforcement Learning and Test-Time Search
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912698962280448 |
|---|---|
| author | Sokota, Samuel Vinitsky, Eugene Hu, Hengyuan Kolter, J. Zico Farina, Gabriele |
| author_facet | Sokota, Samuel Vinitsky, Eugene Hu, Hengyuan Kolter, J. Zico Farina, Gabriele |
| contents | Few classical games have been regarded as such significant benchmarks of artificial intelligence as to have justified training costs in the millions of dollars. Among these, Stratego -- a board wargame exemplifying the challenge of strategic decision making under massive amounts of hidden information -- stands apart as a case where such efforts failed to produce performance at the level of top humans. This work establishes a step change in both performance and cost for Stratego, showing that it is now possible not only to reach the level of top humans, but to achieve vastly superhuman level -- and that doing so requires not an industrial budget, but merely a few thousand dollars. We achieved this result by developing general approaches for self-play reinforcement learning and test-time search under imperfect information. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_07312 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Superhuman AI for Stratego Using Self-Play Reinforcement Learning and Test-Time Search Sokota, Samuel Vinitsky, Eugene Hu, Hengyuan Kolter, J. Zico Farina, Gabriele Machine Learning Artificial Intelligence Few classical games have been regarded as such significant benchmarks of artificial intelligence as to have justified training costs in the millions of dollars. Among these, Stratego -- a board wargame exemplifying the challenge of strategic decision making under massive amounts of hidden information -- stands apart as a case where such efforts failed to produce performance at the level of top humans. This work establishes a step change in both performance and cost for Stratego, showing that it is now possible not only to reach the level of top humans, but to achieve vastly superhuman level -- and that doing so requires not an industrial budget, but merely a few thousand dollars. We achieved this result by developing general approaches for self-play reinforcement learning and test-time search under imperfect information. |
| title | Superhuman AI for Stratego Using Self-Play Reinforcement Learning and Test-Time Search |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2511.07312 |