Tractable Equilibrium Computation in Markov Games through Risk Aversion

arXiv:2406.14156v1 Announce Type: cross
Abstract: A significant roadblock to the development of principled multi-agent reinforcement learning is the fact that desired solution concepts like Nash equilibria may be intractable to compute. To overcome this obstacle, we take inspiration from behavioral economics and show that -- by imbuing agents with important features of human decision-making like risk aversion and bounded rationality -- a class of risk-averse quantal response equilibria (RQE) become tractable to compute in all $n$-player matrix and finite-horizon Markov …

abstract agent agents arxiv behavioral economics computation compute concepts cs.gt cs.lg cs.ma development economics equilibria equilibrium features games important inspiration markov multi-agent reinforcement reinforcement learning risk show solution through tractable type