affiliation not provided to SSRN
continual learning, Pre-trained Models, Parameter-Efficient Fine-Tuning, expert cooperation