Recall Traces: Backtracking Models for Efficient Reinforcement Learning

TL;DR

This paper introduces Recall Traces, a backtracking model that predicts trajectories leading to high-reward states, enhancing sample efficiency in reinforcement learning by focusing training on relevant high-value experiences.

Contribution

It proposes a novel backtracking model to generate high-reward trajectories, improving reinforcement learning efficiency beyond existing methods.

Findings

Improves sample efficiency in RL tasks

Effective for both on-policy and off-policy algorithms

Demonstrates benefits across multiple environments

Abstract

In many environments only a tiny subset of all states yield high reward. In these cases, few of the interactions with the environment provide a relevant learning signal. Hence, we may want to preferentially train on those high-reward states and the probable trajectories leading to them. To this end, we advocate for the use of a backtracking model that predicts the preceding states that terminate at a given high-reward state. We can train a model which, starting from a high value state (or one that is estimated to have high value), predicts and sample for which the (state, action)-tuples may have led to that high value state. These traces of (state, action) pairs, which we refer to as Recall Traces, sampled from this backtracking model starting from a high value state, are informative as they terminate in good states, and hence we can use these traces to improve a policy. We provide a variational interpretation for this idea and a practical algorithm in which the backtracking model samples from an approximate posterior distribution over trajectories which lead to large rewards. Our method improves the sample efficiency of both on- and off-policy RL algorithms across several environments and tasks.