Understanding LoRA Update Complexity Through Stable Rank

Zaynab Awofeso, Antonio Orvieto, Jonas Geiping

Published: 22 Sept 2025, Last Modified: 03 Jan 2026WiML @ NeurIPS 2025EveryoneRevisionsBibTeXCC BY 4.0
Keywords: Parameter-Efficient Fine-Tuning, Low-Rank Adaptation (LoRA), Stable Rank, Update Complexity, Optimization Dynamics, Large Language Models
Submission Number: 225