AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Scalable Learning to Optimize: A Learned Optimizer Can Train Big Models

Xuxi Chen, Tianlong Chen, Yu Cheng, Weizhu Chen, Ahmed Awadallah, Zhangyang Wang

ECCV 2022 · Architectures and efficiency › RNNs and normalization · approach: Dataset distillation and data-efficient training · 5 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
add tiny-subspace reparameterization to MomentumMomentum, ResNet-18 · CIFAR-10convergence steps: 4316.8 -> 216.8Improved
add learned optimizer SL2O on top of subspace trainingMomentum† ResNet-18 · CIFAR-10convergence steps: 216.8 -> 63.6Improved
add learned optimizer SL2O on top of subspace trainingMomentum† ResNet-18 · CIFAR-100convergence steps: 283.2 -> 61.6Improved
add learned optimizer SL2O on top of subspace trainingMomentum† VGG-16 · CIFAR-100convergence steps: 2104.8 -> 168.2Improved
add SL2O instead of MomentumMomentum ResNet-18 · CIFAR-10test accuracy (%): 93.90 -> 93.61No clear effect
add SL2O instead of Momentum†Momentum† VGG-16 · CIFAR-100test accuracy (%): 70.98 -> 71.02No clear effect
add subspace training to SGDSGD VGG-16 · CIFAR-10test accuracy (%): 92.81 -> 89.03Hurt
add subspace training to SGDSGD ResNet-20 · CIFAR-100test accuracy (%): 63.18 -> 66.64Improved
add SL2O instead of Momentum†Momentum† ResNet-18 · ImageNet (40 epochs)test accuracy (%): 68.83 -> 68.78No clear effect
meta-train SL2O on ResNet-18 instead of on ResNet-20 (ablation)SL2O ResNet-20 · CIFAR-10test accuracy (%): 91.00 -> 90.96No clear effect
switch to Momentum after first 5k steps of SL2OSL2O ViT-B/16 · CIFAR-10test accuracy: +0.12%Improved
add SL2O instead of Momentum† 0.04Momentum† ViT-B/16 · CIFAR-10convergence steps: 602 -> 484Improved

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