RFR: Left/Right brain, the hippocampus and episodic learning in AI/ML

RFR: Left/Right brain, the hippocampus and episodic learning in AI/ML

Proposed by Cerenaut and the Whole Brain Architecture Initiative
(What is a Request for Research/RFR?)

The hippocampus is critical for episodic memory, a key component of intelligence, and a sense of self [1].  There are a number of computational models, but none of them consider the fact that the hippocampus is, like the rest of the brain, divided into Left and Right hemispheres.  Division into Left and Right is poorly understood, but undoubtedly critical, as it is a remarkably conserved feature of all bilaterally symmetric animals on Earth. Previously, on a non-episodic classification problem, we mimicked biological differences between hemispheres in left and right neural networks, and achieved specialization and superior performance that matched behavioral observations [2].  This project asks how specialization can improve episodic or one-shot learning by creating a hippocampal model with left and right neural networks.  This will be a truly novel approach to hippocampal modeling, and will help with one of the biggest mysteries in cognitive science, ‘why are brains divided into left and right?’  It also constitutes a new principle in AI/ML.

Required Knowledge:
Machine Learning, Deep Learning or some knowledge and willingness to learn. Must have Python and some experience with PyTorch or Tensorflow.

[1] A. Baddeley, M. Conway, J. Aggleton, and A. Baddeley, “The concept of episodic memory,” Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, vol. 356, no. 1413, pp. 1345–1350, Sep. 2001, doi: 10.1098/rstb.2001.0957.
[2] C. Rajagopalan, D. Rawlinson, E. Goldberg, and G. Kowadlo, “Deep learning in a bilateral brain with hemispheric specialization,” Sep. 2022, doi: 10.48550/arxiv.2209.06862.

Status: Open

Contact: rfr [at]