AI Paper Atlas

Research map / Continual learning / Online and replay-based

Plug-and-Play Compositionality for Boosting Continual Learning with Foundation Models

Weiduo Liao, Fei Han, Hisao Ishibuchi, Qingfu Zhang, Ying Wei

ICLR 2026 · Continual learning › Online and replay-based · approach: Unified and unsupervised continual frameworks

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 CompSLOT plug-inADAM+adapter · CGQA 10-10 tasksAA: 41.930 -> 49.480Improved
add CompSLOT plug-inCPrompt · CGQA 10-10 tasksAA: 46.753 -> 48.537Improved
add CompSLOT plug-inRanPAC · CGQA 10-10 tasksAA: 65.810 -> 66.753Improved
add CompSLOT plug-inFOSTER* · CGQA 10-10 tasksAA: 60.863 -> 66.290Improved
add CompSLOT plug-inEASE · CGQA 10-10 tasksforgetting FF (lower better): 18.215 -> 22.470Hurt
add CompSLOT plug-inDER* · CGQA 10-10 taskscompositional Hn: 90.119 -> 88.986Mixed
add CompSLOT plug-inCPrompt · CGQA 10-10 taskscompositional Hn: 78.063 -> 79.091Improved
increase parameters to match CompSLOT without concept learning (ablation)RanPAC · CGQA 10-10 tasksAA: 65.810 -> 65.080No clear effect
increase prompt length to match CompSLOT without concept learning (ablation)CPrompt · CGQA 10-10 tasksAA: 46.753 -> 46.300No clear effect
add primitive loss and primitive selection (vs slot averaging) (ablation)CompSLOT on RanPAC · CGQA 10-10 tasksAA: 58.220 -> 66.753Improved
add primitive loss and primitive selection (vs slot averaging) (ablation)CompSLOT on CPrompt · CGQA 10-10 tasksAA: 40.230 -> 48.537Improved
use softmax slot weighting instead of cosine (ablation)CompSLOT on RanPAC · CGQA 10-10 tasksAA: 63.910 -> 66.753Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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