{"id":"arxiv/2609.39367","url":"https://arxiv.org/abs/2609.39367","tags":["stat.ML","cond-mat.stat-mech","cs.LG","math.PR","math.ST"],"score":0,"title":"A Dynamical Theory of LoRA in Continual Learning","domain":"stat.ML","source":"arxiv","authors":["Théo Marchetta","Filippo Alessandroni","Alessandro Breccia","Alessandro Ingrosso","Federica Gerace"],"reviews":0,"subject":"stat.ML","summary":"Despite the widespread use of Low-Rank Adaptation (LoRA), little is known about its dynamics in continual learning and the mechanisms by which low-rank updates affect catastrophic forgetting. We provide an asymptotically exact dynamical characterization of LoRA in a solvable two-task teacher-student model. In the high-dimensional online-learning limit, we derive a closed system of ordinary differential equations for a finite set of macroscopic order parameters, yielding exact expressions for the generalization errors throughout both the initial Task 1 learning phase and the subsequent LoRA…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-09-30","reactions":{}}