[pnm]

I figured that solving Calcudoku puzzles would be an ideal test case for the new & improved math/logic reasoning AI agents, so set about testing a couple of the popular ones:

https://www.calcudoku.org/papers/ai_agents/

Comments welcome

(it also seemed like a test that had to be done now, and not in 6 months for example, given the speed of developments in AI land )

This is my Linkedin post about it: https://www.linkedin.com/posts/patrickm ... 41216-dSRh
(in case you want to spread this to your network )