AI Paper Atlas

Research map / Continual learning / Class-incremental

Representation-Steered Incremental Adapter-Tuning for Class-Incremental Learning with Pre-Trained Models

Jiarui Zhao, Libo Huang, Xiangqi Li, Zhulin An, Chuanguang Yang, Yu Wang, Boyu Diao, Yongjun Xu

CVPR 2026 · Continual learning › Class-incremental · approach: Exemplar-efficient class-incremental learning

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 orthogonal loss to the shared-adapter baseline (ablation)SSIAT-style baseline · ImageNet-R B0I20average accuracy: 83.63 -> 84.15 (avg), 79.38 -> 78.32 (final)Mixed
add L2 alignment loss with residual autoencoder projector (ablation)SSIAT-style baseline · ImageNet-A B0I20final accuracy: 62.43 -> 65.04Improved
combine orthogonal and alignment losses (IRS) (ablation)SSIAT-style baseline · ImageNet-R B0I20final accuracy: 79.38 -> 80.52Improved
keep residual identity path in the projector (ablation)IRS projector · ImageNet-A B0I20final accuracy: 51.48 -> 65.31Improved
add base representation-steering loss L_RS (no warm-up) (ablation)RSIAT with IRS · ImageNet-R B0I20average accuracy: 84.86 -> 85.36Improved
add warm-up schedule for the RS coefficient (ablation)RSIAT · ImageNet-R B0I20average accuracy: 85.36 -> 86.92Improved
add warm-up schedule for the RS coefficient (ablation)RSIAT · ImageNet-A B0I20final accuracy: 65.50 -> 66.23Improved
increase adapter bottleneck from 64 to 256 (dp=384) (ablation)RSIAT · ImageNet-R B0I20average accuracy: 86.92 vs 87.04No clear effect
use RSIAT (exemplar-free) instead of exemplar-based traditional CIL methodsViT-B/16-IN21K · CIFAR100 B0I10final accuracy: 89.26 -> 92.20Improved
replace SSIAT with RSIATViT-B/16-IN21K · ImageNet-A B0I20average accuracy: 70.83 -> 74.89Improved

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