I'd look at the underlying paper, https://www.nature.com/articles/s42256-023-00748-9 https://www.nature.com/articles/s42256-023-00748-9. There's neuroscience re

I'd look at the underlying paper, https://www.nature.com/articles/s42256-023-00748-9 https://www.nature.com/articles/s42256-023-00748-9. There's neuroscience references there (50-52) that support their hypothesis. Overall, the corresponding RNN work is meticulously designed and reported, it was a good read.