Researchers have developed a method to reduce the complexity of quantum circuits required for preparing thermofield double (TFD) states, which is crucial for simulating interacting many-body systems. The approach utilizes an adaptation of the multi-angle quantum approximate optimization algorithm (QAOA), termed ma-QAOA, combined with novel sequential pruning techniques. This reduction is vital for implementing quantum algorithms on noisy processors, where circuit depth is a limiting factor.

The study applies ma-QAOA to the preparation of TFD states in Gaussian and binary Sachdev-Ye-Kitaev (SYK) models, in both dense and sparse configurations. SYK models are of particular interest due to their relevance in the study of quantum gravity and condensed matter. The sequential pruning algorithms, which remove Pauli-string evolutions with small optimized angles and reoptimize the remaining parameters, successfully maintain high fidelity in TFD state preparation while significantly reducing circuit depth, especially at low temperatures.

Results show that ma-QAOA prepares target TFD states with high fidelity. For the specific case of the sparse binary N=10 SYK model at β=10, 88.8%–92.1% of nonlocal Pauli-string evolutions were removed while maintaining an average fidelity of approximately 95%. Post-reoptimization of costs after pruning further improved fidelity. This advance is an important step towards efficient quantum simulations of complex systems on noisy platforms, and extensions for quantum-classical hybrid implementations are proposed.