GenQAS Framework Uses Generative Replay for Quantum Architecture Search

GenQAS mitigates sample starvation in quantum circuit search by combining matrix product state warm-starts with a learned local transition model. This tensor network-guided RL framework generates synthetic circuit transitions to improve training signals in shrinking state spaces.