I had a similar idea to convert LLM inputs into Lisp language. I was pleased to see the paper takes this approach.
The underlying idea, though, has been done before. Basically we know that LLMs can do better if instead of solving a problem on their own, they write a program that solves it (look up PAL).
Comments
I had a similar idea to convert LLM inputs into Lisp language. I was pleased to see the paper takes this approach.
The underlying idea, though, has been done before. Basically we know that LLMs can do better if instead of solving a problem on their own, they write a program that solves it (look up PAL).
Here’s an example of a question-and-answer augmentation written in TypeScript that uses the exact approach used by PAL, et al:
https://github.com/williamcotton/transynthetical-engine