Formalizing the Problem of Side Effect RegularizationDownload PDF

Published: 05 Dec 2022, Last Modified: 28 May 2024MLSW2022Readers: Everyone
Abstract: AI objectives are often hard to specify properly. Some approaches tackle this problem by regularizing the AI’s side effects: Agents must weigh off “how much of a mess they make” with an imperfectly specified proxy objective. We propose a formal criterion for side effect regularization via the assistance game framework [Shah et al., 2021]. In these games, the agent solves a partially observable Markov decision process (POMDP) representing its uncertainty about the objective function it should optimize. We consider the setting where the true objective is revealed to the agent at a later time step. We show that this POMDP is solved by trading off the proxy reward with the agent’s ability to achieve a range of future tasks. We empirically demonstrate the reasonableness of our problem formalization via ground-truth evaluation in two gridworld environments.
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 1 code implementation](
1 Reply