Opening review spoke
What is the purpose of supervised fine-tuning in LLM training? — How AI Actually Learns | Unlomi