Thanks. I'm a bit more familiar with neural theorem proving. It's an interesting area. For example, if I could train me a model to speed up (NP-complete) θ-subsumption for very long terms that would be a worthy addition to the purely symbolic toolbox I'm more at home with.

Autoformalization also sounds interesting. I've had some conversations about automatically turning big corpora of natural language text into Prolog with language models, for example. I don't reckon anyone is even researching how to do this with symbolic methods at the moment.

I'll check out AITP. Thanks for the pointers. I'm used to small conferences [and to underciting between disciplines] :)