Efficient training of multiqubit quantum computational sensing protocols

Develop efficient methods for training multiqubit quantum computational sensing protocols that use multiple qubits to encode multiclass outputs, sense signals at multiple spatial locations, or compute more expressive multivariable functions.

Background

The authors propose extending QCS from a single superconducting qubit to multiqubit systems. Multiple qubits could provide enough measurement information to encode multiclass answers, support sensing at different spatial locations, and enable more expressive computations such as multivariable polynomial transformations.

Such extensions introduce a training challenge: the control sequences for multiqubit QCS protocols must be optimized efficiently despite the increased system size and computational complexity. The paper explicitly lists efficient multiqubit-protocol training as an open question.