Spotlight "fine-tuning" Papers

Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation Problem

Maciej Wołczyk, Bartłomiej Cupiał, Mateusz Ostaszewski et al.