AI RESEARCH
Stabilizing LLM Supervised Fine-Tuning via Explicit Distributional Control
arXiv CS.AI
•
ArXi:2605.04468v1 Announce Type: cross Post-