Featured image of post Reward Rational (Implicit) Choice: a unifying formalism for reward learning

Reward Rational (Implicit) Choice: a unifying formalism for reward learning

Hong Jun Jeon; Smitha Milli; Anca Dragan (Neurips 2020)

Authors

Hong Jun Jeon
Department of Computer Science,
Stanford University

Smitha Milli
Department of Computer Science,
UC Berkeley

Anca D. Dragan
Department of Computer Science,
UC Berkeley

Abstract

It is often difficult to hand-specify what the correct reward function is for a task, so researchers have instead aimed to learn reward functions from human behavior or feedback. The types of behavior interpreted as evidence of the reward function have expanded greatly in recent years. We've gone from demonstrations, to comparisons, to reading into the information leaked when the human is pushing the robot away or turning it off. And surely, there is more to come. How will a robot make sense of all these diverse types of behavior? Our key observation is that different types of behavior can be interpreted in a single unifying formalism - as a reward-rational choice that the human is making, often implicitly. We use this formalism to survey prior work through a unifying lens, and discuss its potential use as a recipe for interpreting new sources of information that are yet to be uncovered.

Reference

Neurips Link

Submission

Published at Neurips 2020


Theme Stack designed by Jimmy