Towards Reliable, Consistent, and Safe…

It sounds very interesting, Alejandro. A small demo would be extremely useful!

JonathanR

There look to be examples in the repo, that you can play with 🙂

Thanks😃

Yup, just a few tiny examples for now but I'm working on better documentation.

Thanks for the article. Looking forward to installing and experimenting with the framework. Your articles explaining more of the framework - it's design, usage etc., will be incredibly helpful.

Thanks for the feedback :)

JonathanR

This is extremely interesting and something I am also experimenting with at the moment. Will definitely take a look at ARGO

Thanks! Happy to hear your thoughts and suggestions when you give it shot :)

Mikael

Thank you for your text. You argue that reasoning requires structure. Interestingly, there are different approaches to this. Will think more, and share my thoughts :-)

Thanks! I don't know if I'd go as far as to claim all reasoning requires structure. But at least I think is arguable that some forns of reasoning do benefit a lot from structure. Math and logic come to mind as perhaps obvious examples. And even if structure by itself doesn't actually improve the reasoning capabilities of LLMs (however thin and fragile those may be), at least it does help a lot with verification.

David Sutherland

I'm totally on board!