Thank you! Yours is a very well written article as well. I haven’t tried running KANs on many datasets personally so cannot comment on the overfitting part.