Future Directions in QuantumGEP Research and Development*

QuantumGEP (*) is a free and open source computer program for the generation of quantum circuits, circuits that in turn can generate the quantum mechanical ground state of a given molecular or solid-state Hamiltonian. This talk will discuss a roadmap for improving the convergence of the current implementation (**) by adding a side agent, following the theory behind reinforcement learning with either Q-Learning or prioritized sweeping. Moreover, this talk will discuss plans for better quantum hardware simulation by (i) considering error mitigation in QuantumGEP, and (ii) simulating noisy gates by using a density matrix evaluator instead of a vector-based one.

(*) https://dl.acm.org/doi/10.1145/3617691

(**) https://code.ornl.gov/gonzalo_3/evendim