Replay-buffer engineering for noise-aware quantum circuit optimization.

arXiv:2604.21863v2 Announce Type: replace-cross Abstract: Deep reinforcement learning for quantum circuit optimization faces three bottlenecks: replay buffers that overlook temporal difference (TD) target reliability, curriculum-based architecture search requi