To try everything Brilliant has to offer—free—for a full 30 days, visit http://brilliant.org/ArtemKirsanov/
The first 200 of you will get 20% off Brilliant’s annual premium subscription.

My name is Artem, I'm a computational neuroscience student and researcher. In this video we discuss the TolmanEichenbaum Machine – a computational model of a hippocampal formation, which unifies memory and spatial navigation under a common framework.

Patreon:   / artemkirsanov  
Twitter:   / artemkrsv  

OUTLINE:
00:00 Introduction
01:13 Motivation: Agents, Rewards and Actions
03:17 Prediction Problem
05:58 Model architecture
06:46 Position module
07:40 Memory module
08:57 Running TEM stepbystep
11:37 Model performance
13:33 Cellular representations
17:48 TEM predicts remapping laws
19:37 Recap and Acknowledgments
20:53 TEM as a Transformer network
21:55 Brilliant
23:19 Outro

REFERENCES:
1. Whittington, J. C. R. et al. The TolmanEichenbaum Machine: Unifying Space and Relational Memory through Generalization in the Hippocampal Formation. Cell 183, 12491263.e23 (2020).
2. Whittington, J. C. R., Warren, J. & Behrens, T. E. J. Relating transformers to models and neural representations of the hippocampal formation. Preprint at http://arxiv.org/abs/2112.04035 (2022).
3. Whittington, J. C. R., McCaffary, D., Bakermans, J. J. W. & Behrens, T. E. J. How to build a cognitive map. Nat Neurosci 25, 1257–1272 (2022).

CREDITS:
Icons by biorender.com and freepik.com

Brain 3D models were created with Blender software using publicly available BrainGlobe atlases (https://brainglobe.info/atlasapi)

Animations were made using opensource Python packages Matplotlib and RatInABox ( https://github.com/TomGeorge1234/RatI... )

Rat free 3D model: https://skfb.ly/oEq7y

This video was sponsored by Brilliant