CalebMaresca comments on SAE regularization produces more interpretable models

[CalebMaresca]

Thanks, I’ll take a look!