Abstract
Quantum reinforcement learning (QRL) has emerged as a framework to solvesequential decision-making tasks, showcasing empirical quantum advantages. Anotable development is through quantum recurrent neural networks (QRNNs) formemory-intensive tasks such as partially observable environments. However, QRLmodels incorporating QRNN encounter challenges such as inefficient training ofQRL with QRNN, given that the computation of gradients in QRNN is bothcomputationally expensive and time-consuming. This work presents a novelapproach to address this challenge by constructing QRL agents utilizingQRNN-based reservoirs, specifically employing quantum long short-term memory(QLSTM). QLSTM parameters are randomly initialized and fixed without training.The model is trained using the asynchronous advantage actor-aritic (A3C)algorithm. Through numerical simulations, we validate the efficacy of ourQLSTM-Reservoir RL framework. Its performance is assessed on standardbenchmarks, demonstrating comparable results to a fully trained QLSTM RL modelwith identical architecture and training settings.