Training Language Models for Social Deduction with Multi-Agent RLsocialdeductionllm.github.io3 points·jonbaer··0 commentsOpen articleSaveView on HN