Inference-Aware Fine-Tuning for Best-of-N Sampling in Large Language Models

62 points | by mfiguiere 14 hours ago

8 comments