AI RESEARCH
Reverse Distillation: Consistently Scaling Protein Language Model Representations
arXiv CS.LG
•
ArXi:2603.07710v1 Announce Type: new Unlike the predictable scaling laws in natural language processing and computer vision, protein language models (PLMs) scale poorly: for many tasks, models within the same family plateau or even decrease in performance, with mid-sized models often outperforming the largest in the family. We