Phys.org April 21, 2023
In their previous work an international team of researchers (Australia, Japan, USA – NIST) showed how nanotechnology could be used to build a brain-inspired electrical device with neural network-like circuitry and synapse-like signaling. Their current work paves the way towards replicating brain-like learning and memory in non-biological hardware systems and suggests that the underlying nature of brain-like intelligence may be physical. They implemented task variations inspired by the n-back task (a common memory task used in human psychology experiments) in a nanowire network (NWN) device and applied external feedback to emulate brain-like supervised and reinforcement learning. They found that NWNs retained information in working memory to at least 7 steps back, remarkably like the originally proposed “seven plus or minus two” rule for human subjects. Simulations elucidated how synapse-like NWN junction plasticity depended on previous synaptic modifications, analogous to “synaptic metaplasticity” in the brain, and how memory is consolidated via strengthening and pruning of synaptic conductance pathways… read more. Open Access TECHNICAL ARTICLE
Photograph of nanowire network (left), network’s pathways changing and strengthening (right). Credit: Alon Loeffler.