Fig. 1: HyArch PIC and PIC-RL concepts.
From: High-efficiency reinforcement learning with hybrid architecture photonic integrated circuit
a Schematic of the proposed hybrid architecture PIC (HyArch PIC), comprising a unitary MZI mesh module and three identical parallel OCTOPUS modules. b PIC-assisted reinforcement learning (PIC-RL) leverages PIC for efficient simulation of the agent-environment interaction in the RL algorithm.
