Escaping barren plateaus using entanglement skipping layered ansatz

Thanks to their resilience to noise and wide range of possible applications, Variational Quantum Algorithms (VQA) are among the most useful resources for near-term quantum computing. However, in order to achieve the best performances, there are some issues that need to be addressed. In this talk, I will recap the hybrid quantum-classical structure of VQA, the barren plateau problem and how it connects to the properties of the ansatz. Finally, I will give some possible ways to mitigate the issue and improve trainability of an ansatz by introducing “skip connections” similar to that of classical ResNets in the quantum model.