Category: Supervised Fine-Tuning

• SkillGym: Turning External Skills from Manuals into Executable Supervision So Models Truly Learn Operational Workflows
• ACL 2026 | Why Does SFT Always Fail to Learn? Not Every SFT Failure Needs More Epochs! Five "Surgical Scalpels" to Fix SFT
• LLMs Dominate Math Boards, Yet Forget How to Chat? CMU et al. Reveal Striking Differences Between SFT and RL!