Automated Gadget Discovery in Science

TL;DR

This paper introduces a method to interpret reinforcement learning agents in scientific applications by extracting and clustering frequent subroutines, called gadgets, to understand their learned behaviors.

Contribution

It presents a novel post-hoc analysis technique using sequence mining and clustering to identify meaningful subroutines in RL policies across different scientific domains.

Findings

Gadgets correspond to real experimental setups in quantum optics.

The method successfully identifies quantum information processing subroutines.

Applicable to various agent architectures and environments.

Abstract

In recent years, reinforcement learning (RL) has become increasingly successful in its application to science and the process of scientific discovery in general. However, while RL algorithms learn to solve increasingly complex problems, interpreting the solutions they provide becomes ever more challenging. In this work, we gain insights into an RL agent's learned behavior through a post-hoc analysis based on sequence mining and clustering. Specifically, frequent and compact subroutines, used by the agent to solve a given task, are distilled as gadgets and then grouped by various metrics. This process of gadget discovery develops in three stages: First, we use an RL agent to generate data, then, we employ a mining algorithm to extract gadgets and finally, the obtained gadgets are grouped by a density-based clustering algorithm. We demonstrate our method by applying it to two quantum-inspired RL environments. First, we consider simulated quantum optics experiments for the design of high-dimensional multipartite entangled states where the algorithm finds gadgets that correspond to modern interferometer setups. Second, we consider a circuit-based quantum computing environment where the algorithm discovers various gadgets for quantum information processing, such as quantum teleportation. This approach for analyzing the policy of a learned agent is agent and environment agnostic and can yield interesting insights into any agent's policy.