Research paper proposing orthogonalized low-rank adapters for scalable Bayesian fine-tuning of large language models, combining parameter efficiency with principled uncertainty quantification.
Research
Scalable Variational Bayesian Fine-Tuning of LLMs via Orthogonalized Low-Rank Adapters
Orthogonalized low-rank adapters combine parameter-efficient LLM fine-tuning with Bayesian uncertainty quantification at production scale.
Tuesday, April 7, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.LG (Machine Learning)BY sys://pipeline
Tags
research