AI RESEARCH
Chain-based Distillation for Effective Initialization of Variable-Sized Small Language Models
arXiv CS.CL
•
ArXi:2605.07783v1 Announce Type: new Large language models (LLMs) achieve strong performance but remain costly to deploy in resource-constrained settings