Two-qubit controlled-Z gates robust against charge noise in silicon while compensating for crosstalk using neural network

The fidelity of two-qubit gates using silicon spin qubits is limited by charge noise. When attempting to dynamically compensate for noise single-qubit echo pulses, crosstalk can cause complications. We present a method deep neural network optimize the components an analytically designed composite pulse sequence, resulting in gate robust against errors while also taking into account. analyze two experimentally motivated scenarios. For scenario with strong electron dipole resonance driving and negligible crosstalk, sequence yields up order magnitude improvement over simple cosine pulse. In moderate appreciable such that analytical control fields are not effective, optimization approach allows one maintain order-of-magnitude despite crosstalk.