AI RESEARCH
Not How Many, But Which: Parameter Placement in Low-Rank Adaptation
arXiv CS.AI
•
ArXi:2605.12207v1 Announce Type: cross We study the \textit{parameter placement problem}: given a fixed budget of $k$ trainable entries within the B matrix of a LoRA adapter (A frozen), does the choice of which $k$ matter? Under supervised fine-tuning, random and informed subsets achieve comparable performance. Under GRPO on base models, random placement fails to improve over the base model, while gradient-informed placement recovers standard LoRA accuracy.