Overview
This work introduces machine learning to quantum-state magic estimation. A deep-learning model learns the nonlinear relationship between local measurement features and the magic of ground states and random states.
My Contributions
- Developed a prediction model that estimates quantum-state magic without directly evaluating complex magic measures.
- Achieved relative accuracy above 99%.
- Used tensor networks to simulate systems of tens of qubits and demonstrated size extrapolation from small training systems to larger quantum systems without additional training.