What question did this study set out to answer?

February 16, 20260 citationsOpen Access

Computational Verification of Amartya Sen's Optimal Rationality via Multi-Agent Reinforcement Learning with Meta-Ranking

Puntos clave

To explore Amartya Sen's Meta-Ranking theory within a multi-agent reinforcement learning framework.
Formalization of Meta-Ranking theory in a MARL framework
Large-scale simulations in a Tragedy of the Commons setting
Testing various SVO conditions
Employing causal inference techniques like HAC Robust SE
Identified dynamic moral commitment as crucial for resolving social dilemmas
Found emergent role specialization in simulations
Proved that only Situational Commitment is an Evolutionarily Stable Strategy

Resumen

This study formalizes Amartya Sen's Meta-Ranking theory within a Multi-Agent Reinforcement Learning (MARL) framework, demonstrating that dynamic moral commitment contingent on resource states—rather than simple social preference injection—is the key mechanism for resolving social dilemmas. Through large-scale simulations in a Tragedy of the Commons environment with 7 SVO conditions and rigorous causal inference (HAC Robust SE, p=0.0023), we discover emergent role specialization and prove that only "Situational Commitment" survives as an Evolutionarily Stable Strategy (ESS). Code: https://github.com/Yesol-Pilot/EthicaAI

Preguntar a la IA

Me gusta

Guardar

Ver artículo completo