Deep reinforcement learning for near-deterministic preparation of cubic- and quartic-phase gates in photonic quantum computing

2025-06-09

AI Summary

Non-Gaussian operations—specifically cubic-phase states and quartic-phase gates—are essential for universal continuous-variable photonic quantum computing, yet their preparation suffers from low success probabilities and reliance on gate decomposition schemes. Method: We propose an end-to-end optical circuit optimization framework based on deep reinforcement learning (policy gradient methods), utilizing only photon-number-resolving measurements for real-time feedback control—eliminating the need for intermediate cubic-phase gate decompositions. Contribution/Results: Our approach directly synthesizes high-fidelity quartic-phase gates without decomposition; moreover, we demonstrate for the first time that a single non-Gaussian resource—photon-number-resolving measurement—enables efficient preparation of both cubic-phase states (average success probability 96%) and quartic-phase gates. Experimentally, the synthesized quartic gates achieve significantly higher fidelity than those obtained via conventional decomposition-based methods. This work overcomes the gate-decomposition bottleneck and establishes a new paradigm for scalable, high-efficiency continuous-variable quantum computation.

Technology Category

Machine Learning: Quantum Machine LearningComputer Vision: Learning & Optimization for CVSearch and Optimization: Learning to Search

Application Category

Economics, Online Markets and Human Computation: LLM based quality controls for crowd workSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingSecurity and Privacy: Large-scale security measurements

Abstract

Cubic-phase states are a sufficient resource for universal quantum computing over continuous variables. We present results from numerical experiments in which deep neural networks are trained via reinforcement learning to control a quantum optical circuit for generating cubic-phase states, with an average success rate of 96%. The only non-Gaussian resource required is photon-number-resolving measurements. We also show that the exact same resources enable the direct generation of a quartic-phase gate, with no need for a cubic gate decomposition.

Problem

• Generating cubic-phase states for photonic quantum computing
• Achieving high success rate with deep reinforcement learning
• Direct generation of quartic-phase gate without decomposition

Innovation

• Deep reinforcement learning controls quantum optical circuits
• Photon-number-resolving measurements enable non-Gaussian resources
• Direct quartic-phase gate generation without cubic decomposition