Inferring Implicit Goals Across Differing Task Models

TL;DR

This paper proposes a method to infer implicit user subgoals in tasks modeled as MDPs by identifying bottleneck states and querying efficiently, improving alignment with user expectations.

Contribution

It introduces a novel approach to detect and query implicit subgoals in differing task models, enhancing value alignment in AI systems.

Findings

Effective identification of implicit subgoals in various tasks.

Minimal querying strategy achieves accurate goal inference.

Improved alignment between agent behavior and user expectations.

Abstract

One of the significant challenges to generating value-aligned behavior is to not only account for the specified user objectives but also any implicit or unspecified user requirements. The existence of such implicit requirements could be particularly common in settings where the user's understanding of the task model may differ from the agent's estimate of the model. Under this scenario, the user may incorrectly expect some agent behavior to be inevitable or guaranteed. This paper addresses such expectation mismatch in the presence of differing models by capturing the possibility of unspecified user subgoal in the context of a task captured as a Markov Decision Process (MDP) and querying for it as required. Our method identifies bottleneck states and uses them as candidates for potential implicit subgoals. We then introduce a querying strategy that will generate the minimal number of queries required to identify a policy guaranteed to achieve the underlying goal. Our empirical evaluations demonstrate the effectiveness of our approach in inferring and achieving unstated goals across various tasks.