Fig. 2: Schematic depiction of the recursive optimization process, demonstrating the incremental increase in circuit depth p and system size n, and potential pathways for parameter transfer optimization.
From: Unveiling quantum phase transitions from traps in variational quantum algorithms
Although setting p = n/2 allows for the exact preparation of ground states as explored in63, our study focuses on the dynamics of locally trapped states under conditions where n ≥ 2p + 2, thus examining the effects of finite circuit depth. Importantly, for any given p, systems with n ≥ 2p + 2 can share the same optimized parameters. In our simulations, we specifically adopt n = 2p + 2 for optimization, as highlighted by the dotted dark green arrow, to maintain a consistent approach across different system sizes.
