Comment by hendler
13 hours ago
I also wrote wrote an LLM to Prolog interpreter for a hackathon called "Logical". With a few hours effort I'm sure it could be improved.
https://github.com/Hendler/logical
I think while LLMs may approach completeness here, it's good to have an interpretable system to audit/verify and reproduce results.
This is really cool!