Learning Arbitrary Lindbladians with Quantum Error Correction
This research develops an ansatz-free method to learn quantum system generators, like Hamiltonians and dissipators, from data without knowing their structure beforehand. It uses recursive Quantum Error Correction (QEC) to sequentially suppress strong unknown terms while preserving sensitivity to weaker ones. The method achieves Heisenberg scaling for the Hamiltonian part and standard-quantum-limited scaling for the full evolution.