Fig. 4: Summary multiple injection QELMs.
From: Potential and limitations of quantum extreme learning machines

The reservoir interacts with multiple copies of the same quantum state ρ and progressively acquires information about it. The measurement is performed on the final reservoir state ηn.