Yes, that’s my understanding too. GAMs have structural limitations that harm flexibility and help interpretability. One benefit of Bayesian models that they wo

Yes, that’s my understanding too. GAMs have structural limitations that harm flexibility and help interpretability. One benefit of Bayesian models that they work with relatively little data - and generally provide greater uncertainty in those cases. Do you happen to know of some DL frameworks that behave similarly? I’m eager to learn.