AI Paper Atlas

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

Null-Space Filtering for Data-free Continual Model Merging: Preserving Transparency, Promoting Fidelity

Zihuan Qiu, Lei Wang, Yang Cao, Runtong ZHANG, Bing Su, Yi Xu, Fanman Meng, Linfeng Xu, Qingbo Wu, Hongliang Li

ICLR 2026 · Continual learning › Continual fine-tuning of LLMs · approach: Synthetic data and merging for CLIP

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
add null-space projection to naive task-vector merging (ablation)naive merging · ViT-B/32, 8 tasksACC: 62.1 -> 80.0Improved
add null-space projection to naive merging (ablation)naive merging · ViT-B/32, 20 tasksACC: 34.3 -> 67.0Improved
add null-space projection to naive merging (ablation)naive merging · ViT-B/32, 8 tasksBWT: -18.5 -> -1.7Improved
add data-free low-rank objective to null-space filter (ablation)NUFILT · ViT-B/32, 8 tasksACC: 80.0 -> 83.6Improved
add data-free low-rank objective to null-space filter (ablation)NUFILT · ViT-B/32, 20 tasksACC: 67.0 -> 71.0Improved
add data-free low-rank objective to null-space filter (ablation)NUFILT · ViT-B/32, 8 tasksBWT: -1.7 -> -2.7Hurt
add null-space filter to data-free objective (ablation)NUFILT · ViT-B/32, 8 tasksACC: 75.8 -> 83.6Improved
add null-space filter to data-free objective (ablation)NUFILT · ViT-B/32, 20 tasksBWT: -20.6 -> -8.9Improved
add data-free objective alone to naive merging (ablation)naive merging · ViT-B/32, 8 tasksACC: 62.1 -> 75.8Improved
increase null-space rank r_p excessively (ablation)NUFILT · ViT-B/32ACCHurt
increase low-rank adaptation rank r_l (ablation)NUFILT · ViT-B/32ACCImproved
use NUFILT instead of OPCMOPCM · ViT-L/14, 20 tasksACC: 76.0 -> 84.7Improved

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