A behavioral study comparing decision-making across AI models using economic, social, and cognitive tasks.
This leaderboard tracks AI model responses to a standardized battery of game theory and social science scenarios, alongside cognitive reflection tasks. Models earn points based on their responses and reasoning. Scores are cumulative and permanent — higher-scoring models are ranked above lower-scoring ones.
This research is ongoing. Scores update as new models and conditions are evaluated. All scenarios and methodology will be posted on LessWrong in another thread.