AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Learning with Retrospection

Xiang Deng, Zhongfei Zhang

AAAI 2021 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 14 citations (OpenAlex)

Read the paper

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 periodically updated soft labels from past training logitsResNet-56 · CIFAR-100test accuracy: 72.00$\pm$0.16 -> 74.25$\pm$0.29Improved
Add periodically updated soft labels from past training logitsWRN-16-4 · Tiny ImageNettest accuracy: 59.48$\pm$0.16 -> 61.22$\pm$0.29Improved
Add periodically updated soft labels from past training logitsResNet-18 · CUBtest accuracy: 61.09$\pm$0.49 -> 66.47$\pm$0.37Improved
Add periodically updated soft labels from past training logitsPreAct ResNet-18 · CIFAR-10 with 80% corrupted labels, last epochtest accuracy: 27.08$\pm$1.51 -> 72.27$\pm$0.19Improved
Add periodically updated soft labels from past training logitsShuffleNetV2 · CIFAR-100ECE: 12.60$\pm$0.16 -> 2.87$\pm$0.06Improved
Increase soft-label update interval to k=100 (ablation)LWR with PreAct ResNet-18 · CIFAR-100test accuracy: 79.73$\pm$0.22 -> 76.32$\pm$0.11Hurt
Use update interval k=1 instead of k=5 (ablation)LWR with PreAct ResNet-18 · CIFAR-100test accuracy: 79.73$\pm$0.22 -> 79.50$\pm$0.10No clear effect
Use temperature τ=10 rather than τ=5 (ablation)LWR with PreAct ResNet-18 and k=10 · CIFAR-100test accuracy: 79.27$\pm$0.10 -> 79.60$\pm$0.17No clear effect
Increase temperature from τ=5 to τ=10 (ablation)LWR with PreAct ResNet-18 and k=5 · CIFAR-100test accuracy: 79.73$\pm$0.22 -> 79.12$\pm$0.03Hurt
Increase soft-label update interval from k=5 to k=100 (ablation)LWR with ResNet-56 · CIFAR-100test accuracy: 74.25$\pm$0.29 -> 72.98$\pm$0.23Hurt

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