AI Paper Atlas

Research map / Continual learning / Online and replay-based

SplitLoRA: Balancing Stability and Plasticity in Continual Learning Through Gradient Space Splitting

Haomiao Qiu, Miao Zhang, Ziyue Qiao, Weili Guan, Min Zhang, Liqiang Nie

ICLR 2026 · Continual learning › Online and replay-based · approach: Unified and unsupervised continual frameworks · 5 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
project A_t onto minor subspace (SplitLoRA init) instead of random init (ablation)SplitLoRA · ImageNet-R 20-taskFAA: 72.30 -> 78.82Improved
project A_t onto minor subspace instead of InfLoRA initialization (ablation)SplitLoRA · ImageNet-R 20-taskFAA: 73.81 -> 78.82Improved
choose per-module subspace size by stability-plasticity objective instead of fixed thresholdSplitLoRA vs InfLoRA · ImageNet-R 10-taskFAA: 78.10 -> 81.00Improved
vary alpha from 20 to 1, 5, 10 or 30 (ablation)SplitLoRA · ImageNet-RFAA: 80.89 - 81.38 (10-task)No clear effect
increase alpha (ablation)SplitLoRA · ImageNet-Rforgetting / plasticityMixed
swap ViT pretraining to iBOT-1KSplitLoRA vs VPT-NSP2 · ImageNet-R 10-taskFAA: 73.85 -> 74.58Improved
swap ViT pretraining to DINO-1KSplitLoRA vs VPT-NSP2 · ImageNet-R 10-taskFAA: 69.45 -> 70.49Improved
use SplitLoRA instead of InfLoRA on DomainNetSplitLoRA · DomainNet 5-taskFAA: 78.26 -> 84.31Improved
use SplitLoRA instead of InfLoRA on efficiencySplitLoRA · ImageNet-R 10-taskextra forward passes / time: 2/task -> 1/task; 1h 48m -> 1h 43mImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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