Reinforcement Learning for Adaptive Composition of Quantum Circuit Optimisation Passes

TL;DR

This paper introduces a reinforcement learning approach to automatically compose quantum circuit optimisation pass sequences, significantly improving two-qubit gate reduction over default sequences.

Contribution

It presents a novel RL-based method for tailoring quantum circuit optimisation sequences, outperforming default pass sequences in gate reduction.

Findings

RL agent achieves 57.7% two-qubit gate reduction

Outperforms default pass sequences in diverse circuits

Demonstrates potential for circuit-specific optimisation

Abstract

Many quantum software development kits provide a suite of circuit optimisation passes. These passes have been highly optimised and tested in isolation. However, the order in which they are applied is left to the user, or else defined in general-purpose default pass sequences. While general-purpose sequences miss opportunities for optimisation which are particular to individual circuits, designing pass sequences bespoke to particular circuits requires exceptional knowledge about quantum circuit design and optimisation. Here we propose and demonstrate training a reinforcement learning agent to compose optimisation-pass sequences. In particular the agent's action space consists of passes for two-qubit gate count reduction used in default PyTKET pass sequences. For the circuits in our diverse test set, the (mean, median) fraction of two-qubit gates removed by the agent is $(57.7\%, \ 56.7 \%)$, compared to $(41.8 \%, \ 50.0 \%)$ for the next best default pass sequence.