A Novel Modular Neural Architecture for Rule-Based and Similarity-Based Reasoning

Bogacz R
Giraud-Carrier C
Scientific Abstract

Hybrid connectionist symbolic systems have been the subject of much recent research in AI. By focusing on the implementation of high- level human cognitive processes (e.g., rule-based inference) on low-level, brain-like structures (e.g., neural networks), hybrid systems inherit both the eciency of connectionism and the comprehensibility of symbolism. This paper presents the Basic Reasoning Applicator Implemented as a Neural Network (BRAINN). Inspired by the columnar organisation of the human neocortex, BRAINN's architecture consists of a large hexagonal network of Hopeld nets, which encodes and processes knowledge from both rules and relations. BRAINN supports both rule-based reasoning and similarity-based reasoning. Empirical results demonstrate promise.

Citation
2000.Hybrid Neural Systems, LNAI 1778, pp. 63–77
More Like This
Publication
Barron HC, Reeve HM, Koolschijn RS, Perestenko PV, Shpektor A, Nili H, Rothaermel R, Campo-Urriza N, O'Reilly JX, Bannerman DM, Behrens TE, Dupret D
2020. Cell, 183(1):228-243.e21.
Unit Publication
Publication
Tan H, Debarros J, He S, Pogosyan A, Aziz TZ, Huang Y, Wang S, Timmermann L, Visser-Vandewalle V, Pedrosa D, Green AL, Brown P
2019. Brain Stimul., 12(4):858-867.
Unit Publication
Publication
El-Gaby M, Reeve HM, Lopes-Dos-Santos V, Campo-Urriza N, Perestenko PV, Morley A, Strickland LAM, Lukács IP, Paulsen O, Dupret D
2021. Nat Neurosci, 24:694-704.
Unit Publication
Publication
He S, Baig F, Mostofi A, Pogosyan A, Debarros J, Green AL, Aziz TZ, Pereira EA, Brown P, Tan H
2021. Mov Disord, 36(4):863-873.
Unit Publication