lechmazur/elimination_game
A multi-player tournament benchmark that tests LLMs in social reasoning, strategy, and deception. Players engage in public and private conversations, form alliances, and vote to eliminate each other
194 best points · 60 comments · 3 Hacker News posts · 298 GitHub stars · View on GitHub
Hacker News discussions
- Benchmarking LLM social skills with an elimination game (194 points, 60 comments, by colonCapitalDee, 2025-04-04)
- Elimination Game: Multi-Agent LLM Social Reasoning, Strategy, and Deception (5 points, 0 comments, by zone411, 2025-02-25)
- Elimination Game Benchmark: Social Reasoning, Strategy, and Deception in LLMs (2 points, 0 comments, by amichail, 2025-03-03)