Behrens' Bayesian Learner Model : How to share parameters between steps?

I think so. My model looks more “sensible” than Waskom’s model.

And I think my model still catches the volatile of the sequence and the probability of p. So I think I do reproduce the model of Behrens.

I may write a tutorial next week (my supervisor gives me a huge of work on my fmri data, and the DDL is next Monday) and let more people check our model and through.