LLM Adaptation and Post-Training
Keep a pretrained linear weight frozen and learn a small low-rank delta that can be applied or merged for an adapted model.
JavaScript enhances this canonical concept page with formulas, examples, tools, and related routes.
