AI Paper Atlas

Research map / Continual learning / Continual fine-tuning of LLMs

Understanding Catastrophic Forgetting In LoRA via Mean-Field Attention Dynamics

Hugo Koubbi, Louis Hernandez, Matthieu Boussard

ICML 2026 · Continual learning › Continual fine-tuning of LLMs · approach: Low-rank and model-merging adaptation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
increase Frobenius norm of random Gaussian low-rank perturbation added to weights (ablation)Qwen3-0.6B · next-token loss on base datasetperturbed lossHurt
increase rank of random Gaussian perturbation at fixed Frobenius norm (ablation)Qwen3-0.6B · next-token loss on base datasetperturbed lossNo clear effect
vary LoRA rank during actual training (ablation)Qwen3-0.6B LoRA · base-dataset perplexitynon-trained perplexityMixed
continue LoRA fine-tuning (accumulating perturbation)Qwen3-0.6B LoRA · hidden-state drift vs base modelempirical bifurcation depth (layer): 28 -> 16Hurt
low-rank LoRA adapters (r=8, 32) during trainingQwen3-0.6B LoRA · alignment with dominant value direction u1normalized subspace alignmentHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy