AI Paper Atlas

Research map / Continual learning / Domain and prompt-based

Hierarchical Decomposition of Prompt-Based Continual Learning: Rethinking Obscured Sub-optimality

Liyuan Wang, Jingyi Xie, Xingxing Zhang, Mingyi Huang, Hang Su, Jun Zhu

NeurIPS 2023 · Continual learning › Domain and prompt-based · approach: Prompt-based continual learning · 19 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
replace existing prompt-based continual-learning methods with HiDe-PromptViT-B/16 continual learning · Split CIFAR-100 and Split ImageNet-R across supervised and self-supervised pre-trainingFAA: up to 15.01% and 9.61% leadImproved
add prompt ensemble and WTP (ablation)naive task-specific prompt architecture · Split CIFAR-100 and Split ImageNet-R, across pre-training paradigmsFAA: 72.20 -> 75.93; 53.68 -> 60.86Improved
add TII to WTP (ablation)HiDe-Prompt component ablation · Split CIFAR-100, iBOT-1KFAA: 75.93 -> 78.90Improved
add TII to WTP (ablation)HiDe-Prompt component ablation · Split ImageNet-R, iBOT-1KFAA: 60.86 -> 61.22Improved
add TAP to WTP (ablation)HiDe-Prompt component ablation · Split CIFAR-100, iBOT-1KFAA: 75.93 -> 86.04Improved
add TAP to WTP (ablation)HiDe-Prompt component ablation · Split ImageNet-R, iBOT-1KFAA: 60.86 -> 67.13Improved
add both TII and TAP to WTP (ablation)HiDe-Prompt component ablation · Split CIFAR-100, iBOT-1KFAA: 75.93 -> 90.52Improved
add both TII and TAP to WTP (ablation)HiDe-Prompt component ablation · Split ImageNet-R, iBOT-1KFAA: 60.86 -> 70.66Improved
add contrastive regularization (ablation)WTP+TII+TAP · Split CIFAR-100, iBOT-1KFAA: 90.52 -> 93.48Improved
add contrastive regularization (ablation)WTP+TII+TAP · Split ImageNet-R, iBOT-1KFAA: 70.66 -> 71.33Improved
increase contrastive-regularization strength (ablation)WTP with CR · Split CIFAR-100 and Split ImageNet-R, iBOT-1K and DINO-1KFAA: 75.13 -> 71.53; 61.47 -> 59.52; 74.99 -> 70.87; 57.71 -> 56.37Hurt
increase contrastive-regularization strength (ablation)WTP+TII+TAP with CR · Split CIFAR-100 and Split ImageNet-R, iBOT-1K and DINO-1KFAA: 90.15 -> 93.35; 70.97 -> 71.71; 88.97 -> 92.61; 66.88 -> 68.21Improved

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