I don't think that framing this specifically in terms of curiosity is a good idea, as distinct from just detecting useless activity (i.e. lost purposes). Same thing as what JGWeissman said.

For example, when applied to my revisiting of basics of mathematics/logic over the last year or so, it could well give a false positive, since I wasn't particularly curious about any tool/idea I was learning, and I was not trying to solve any particular problem that better knowledge of math would help me with. The goal was simply to obtain better skills, so that thinking about decision theory can become more clear/fruitful, in whatever ways.