AI Paper Atlas

Research map / Continual learning / Class-incremental

The Lie of the Average: How Class Incremental Learning Evaluation Deceives You?

Guannan Lai, Da-Wei Zhou, Xin Yang, Han-Jia Ye

ICLR 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
use EDGE (easy/median/hard sequences by inter-task similarity) instead of 3 random sequencesHide-Prompt · CIFAR-100 (6 classes, 3 tasks)Wasserstein distance: 3.89 -> 1.42Improved
use EDGE instead of random samplingCODA-Prompt · ImageNet-R (6 classes, 3 tasks)Wasserstein distance: 9.85 -> 2.37Improved
use EDGE instead of random samplingEWC · CIFAR-100 (6 classes, 3 tasks)estimated lower bound accuracy (true 12.50): 26.17 -> 12.50Improved
use EDGE instead of random sampling on stable models (EASE, RanPAC)EASE, RanPAC · CIFAR-100 and ImageNet-R (6 classes)JSD / Wasserstein: 0.00 vs 0.00 JSD on CIFAR-100No clear effect
use EDGE instead of random samplingDER · CIFAR-100 (6 classes, 3 tasks)Jensen-Shannon divergence: 0.29 -> 0.31Hurt
vary CLIP text encoder scale and backbone (ResNet vs ViT) in EDGE (ablation)EDGE · Hide-Prompt and TagFex enumerabledistribution estimate closenessNo clear effect
add extreme sequences to the evaluation samplerandom sampling evaluation · theoryrequired number of sequences L: 2x10^4 -> around 50Improved
evaluate on easier class sequence with higher inter-task similarityCIL methods · CIFAR-100 and ImageNet-R enumerableaccuracy vs similarityImproved

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