{"id":"arxiv/2605.10910","url":"https://arxiv.org/abs/2605.10910","tags":["quant-ph","cs.LG"],"score":0,"title":"Equivariant Reinforcement Learning for Clifford Quantum Circuit Synthesis","domain":"quant-ph","source":"arxiv","authors":["Richie Yeung","Aleks Kissinger","Rob Cornish"],"reviews":0,"subject":"quant-ph","summary":"We consider the problem of synthesizing Clifford quantum circuits for devices with all-to-all qubit connectivity. We approach this task as a reinforcement learning problem in which an agent learns to discover a sequence of elementary Clifford gates that reduces a given symplectic matrix representation of a Clifford circuit to the identity. This formulation permits a simple learning curriculum based on random walks from the identity. We introduce a novel neural network architecture that is equivariant to qubit relabelings of the symplectic matrix representation, and which is size-agnostic,…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-05-11","reactions":{}}