AI Paper Atlas

Research map / Continual learning / Online and replay-based

TRGP: Trust Region Gradient Projection for Continual Learning

Sen Lin, Li Yang, Deliang Fan, Junshan Zhang

ICLR 2022 · Continual learning › Online and replay-based · approach: Unified and unsupervised continual frameworks · 22 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
use TRGP instead of GPMcontinual learning · PMNISTACC (%): 93.91 -> 96.34Improved
use TRGP instead of GPMcontinual learning · CIFAR-100 SplitACC (%): 72.48 -> 74.46Improved
use TRGP instead of HATcontinual learning · 5-DatasetACC (%): 91.32 -> 93.56Improved
use TRGP instead of GPMcontinual learning · MiniImageNetACC (%): 60.41 -> 61.78Improved
use TRGP instead of GPMcontinual learning · CIFAR-100 SupACC (%): 57.72 -> 58.25Improved
vary εˡ among 0.2, 0.5, and 0.7 (ablation)TRGP · CIFAR-100 SplitACC: 74.52, 74.46, 74.30No clear effect
vary εˡ among 0.2, 0.5, and 0.7 (ablation)TRGP · 5-DatasetACC: 93.28, 93.56, 93.43No clear effect
use a fixed task-wise trust region across layers instead of layer-wise selection (ablation)TRGP · CIFAR-100 SplitACC: 74.46 -> 73.25Hurt
use a fixed task-wise trust region across layers instead of layer-wise selection (ablation)TRGP · 5-DatasetACC: 93.56 -> 92.85Hurt
select top-2 correlated old tasks instead of top-1 (ablation)TRGP · CIFAR-100 SplitACC: 74.00 -> 74.46Improved
select top-2 correlated old tasks instead of top-1 (ablation)TRGP · 5-DatasetACC: 92.94 -> 93.56Improved

Later papers that compared with this method

10 later accepted papers report TRGP as a baseline in their results tables; 9 of them report a better number on most of the results they share with it.

2 more not listed.

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