Controllable Quantum Memory Capacity in Quantum Reservoir Networks with Tunable partial-SWAPs
Researchers have developed a new mechanism called a tunable partial-SWAP to enhance control over memory capacity in quantum reservoir networks (QRNs). This advancement builds upon existing recurrent QRC architectures, which use multiple registers to create a fading memory, but often lack a clear understanding or direct control over this memory mechanism. The tunable partial-SWAP allows for direct manipulation of how quickly memory dissipates within a QRN implemented on gate-based quantum processing units (QPUs). The effectiveness of this mechanism was validated through experiments using a randomized short-term memory capacity benchmark and the NARMA-5 dataset, with results tested on IBM QPUs. AI
IMPACT Enhances control over quantum memory, potentially improving performance in quantum machine learning applications.