Train AI to adapt like human brains
November 2020
Salk Institute for Biological Studies, La Jolla, USA
The prefrontal cortex (PFC) enables humans’ ability to flexibly adapt to new environments and circumstances. Disruption of this ability is often a hallmark of prefrontal disease. Neural network models have provided tools to study how the PFC stores and uses information, yet the mechanisms underlying how the PFC is able to adapt and learn about new situations without disrupting preexisting knowledge remain unknown. Here a neural network architecture called DynaMoE is used to show how hierarchical gating can naturally support adaptive learning while preserving memories from prior experience. Furthermore, the authors show how damage to the network model recapitulates disorders of the human PFC.
A modeling framework for adaptive lifelong learning with transfer and savings through gating in the prefrontal cortex
Terrence J. Sejnowski, Ben Tsuda
Added on: 12-21-2020
[1] https://www.pnas.org/content/117/47/29872[2] https://www.technologynetworks.com/informatics/news/can-we-train-ai-to-adapt-like-human-brains-344096
