Barren plateaus from learning scramblers with local cost functions

A bstract The existence of barren plateaus has recently revealed new training challenges in quantum machine learning (QML). Uncovering the mechanisms behind is essential understanding scope problems that QML can efficiently tackle. Barren have been shown to exist when global properties random unitaries, which relevant black hole dynamics. Establishing whether local cost functions circumvent these pertinent if we hope apply many-body systems. We prove a no-go theorem showing encounter unitary properties.