S seantey.github.io I used RL fine-tuning to make an LLM generate ugly and unpythonic FizzBuzz code 4 points seanrrr 10 months ago 1 comments
