Hacker Timesnew | past | comments | ask | show | jobs | submitlogin

I wonder if someone has tried to chain ChatGPT's output with static analyzers/compilers/linters in a loop to improve the final response.


https://github.com/newhouseb/clownfish

Structural Alignment: Modifying Transformers (like GPT) to Follow a JSON Schema




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: