AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

Which Model To Transfer? Finding the Needle in the Growing Haystack

Cedric Renggli, André Susano Pinto, Luka Rimanic, Joan Puigcerver, Carlos Riquelme, Ce Zhang, Mario Lučić

CVPR 2022 · Architectures and efficiency › Neural architecture search · approach: AutoML pipeline search · 14 citations (OpenAlex)

Check 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
Combine the top task-agnostic model with the top task-aware linear-evaluation modelModel search across the All, ResNet-50, and Expert pools · VTAB-1Krelative regretImproved
Include the task-agnostic pick alongside the linear-evaluation pickLinear evaluation, ResNet-50 pool · dSpr-Locrelative regret: 36%Improved
Include the task-agnostic pick alongside the linear-evaluation pickLinear evaluation, Expert pool · KITTI-Distrelative regret: 11%Improved
Use linear evaluation rather than task-agnostic rankingModel search, Expert pool · VTAB-1Krelative regretImproved
Use linear evaluation rather than task-agnostic rankingModel search, All pool · VTAB-1Krelative regretMixed
Use linear evaluation rather than task-agnostic rankingModel search, ResNet-50 or Expert pool · Structured datasetsrelative regretMixed
Use kNN instead of linear evaluation as the task-aware proxy (ablation)Model search across model pools · VTAB-1Krelative regretMixed
Use kNN rather than linear evaluation in the hybrid strategy (ablation)Hybrid model search, All pool · VTAB-1Krelative regretImproved
Use kNN rather than linear evaluation in the hybrid strategy (ablation)Hybrid model search, Expert pool · VTAB-1Krelative regretHurt
Replace linear evaluation with early stopping after 10 SGD iterations (ablation)Task-aware model search · VTAB-1Krelative regretNo clear effect
Increase kNN neighborhood size from k=1 to k=3 (ablation)kNN evaluation · VTAB-1KaccuracyNo clear effect
Use ImageNet accuracy ranking instead of task-aware search (ablation)Model search, ImageNet-accuracy pool · VTAB-1Knumber of optimal model picks: 12 out of 19 for B=1; 17 out of 19 for B=2Improved

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