Deep Learning Framework for Artificial General Intelligence

EasyChair Preprint 7921, version 1

Abstract

This paper proposes two Deep Learning (DL) related models, to serve potentially as parts of an AGI agent. The first one is designed bottom-up, i.e. it is mostly based on DL. The second one is a partial AGI model, specifically concerning the thinking process. It is designed top-down, i.e. it is mainly based on cognition and communication. The latter has not yet been fully designed for implementation. It only describes the representation of the data, and its relevance to DL is by being triggered by some Deep Neural Network (DNN).

Keyphrases: Neuro Science, associative thinking, deep learning

@booklet{EasyChair:7921,
  author    = {Shimon Komarovsky},
  title     = {Deep Learning Framework for Artificial General Intelligence},
  howpublished = {EasyChair Preprint 7921},
  year      = {EasyChair, 2022}}