Super human Stratego with RL and test time search | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		Super human Stratego with RL and test time search (arxiv.org)
		2 points by algo_trader 3 months ago \| hide \| past \| favorite \| 1 comment

algo_trader 3 months ago [–]

Only 2000 GPU hours Heavily customized network 95% win rate in recent human tournament sample Several training techniques for evaluation/learning rate

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact