This is a special post for quick takes by leogao. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
idea: maybe ambitious and pragmatic interp can "meet in the middle" by explaining larger and larger models on the one hand, and reducing the problem of understanding a large model to the problem of understanding some smaller explainer model on the other hand.
I'm confused. What are some examples of pragmatic interp creating smaller explainer models which we can use ambitious interp methods on? Why do we need pragmatic interp for this purpose instead of e.g. distillation?