Cooperative Inverse Reinforcement Learning

TL;DR

This paper introduces cooperative inverse reinforcement learning (CIRL), a framework where humans and robots collaborate with the robot learning human values through interaction, leading to more effective value alignment strategies.

Contribution

It formalizes the value alignment problem as CIRL, demonstrating its connection to POMDPs and developing an approximate algorithm for solving CIRL games.

Findings

Optimal CIRL solutions involve active teaching and communication.

Computing optimal policies reduces to solving a POMDP.

Optimality in isolation is suboptimal in CIRL.

Abstract

For an autonomous system to be helpful to humans and to pose no unwarranted risks, it needs to align its values with those of the humans in its environment in such a way that its actions contribute to the maximization of value for the humans. We propose a formal definition of the value alignment problem as cooperative inverse reinforcement learning (CIRL). A CIRL problem is a cooperative, partial-information game with two agents, human and robot; both are rewarded according to the human's reward function, but the robot does not initially know what this is. In contrast to classical IRL, where the human is assumed to act optimally in isolation, optimal CIRL solutions produce behaviors such as active teaching, active learning, and communicative actions that are more effective in achieving value alignment. We show that computing optimal joint policies in CIRL games can be reduced to solving a POMDP, prove that optimality in isolation is suboptimal in CIRL, and derive an approximate CIRL algorithm.