category theory?

[Ross Gayler]

In the Q&A of Mike Heddes' talk, Pentti Kanerva mentioned getting mathematicians to take an interest in VSA/HDC.

Approaching that idea from the other direction, I note that some deep-learning/machine-learning people are starting to apply category theory to their work. Given that VSA/HDC are implementable as neural networks, I would hope that they would fit in. Given that VSA/HDC have a much more algebraic feel to them than other neural networks it's possible they might be better suited to a category theoretic treatment.

If that sounds interesting to you this set of lectures on applied category theory in deep learning might be of some use to you:

https://cats.for.ai/program/

Cheers,

Ross