AI RESEARCH
Team-Based Self-Play With Dual Adaptive Weighting for Fine-Tuning LLMs
arXiv CS.AI
•
ArXi:2605.09922v1 Announce Type: cross While recent self-