AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Boost Neural Networks by Checkpoints

Feng Wang, Guoyizhe Wei, Qiao Liu, Jinxiang Ou, Xian Wei, Hairong Lv

NeurIPS 2021 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 0 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 adaptive sample reweighting to train and ensemble checkpoints instead of the strongest checkpoint-ensemble baselineResNet-110 · Cifar-100error rate (%): 24.27 ± 0.26 -> 23.51 ± 0.20Improved
Use CBNN rather than the strongest checkpoint-ensemble baselineDenseNet-40 · Tiny-ImageNeterror rate (%): 36.35 ± 0.30 -> 34.28 ± 0.20Improved
Use CBNN rather than the strongest checkpoint-ensemble baselineEfficientNet-B0 · Cifar-100error rate (%): 19.58 ± 0.24 -> 18.06 ± 0.20Improved
Use CBNN rather than the strongest checkpoint-ensemble baselineEfficientNet-B3 · Cifar-10error rate (%): 3.27 -> 3.28Mixed
Use CBNN rather than a single modelEfficientNet-B0 · Cifar-100 (imbalanced)error rate (%): 29.19 -> 24.39Improved
Use CBNN rather than a single modelEfficientNet-B3 · Tiny-ImageNet (imbalanced)error rate (%): 43.34 -> 39.02Improved
Use CBNN instead of a single model to reach the 70% accuracy targetEfficientNet-B0 · Cifar-100GPU hours: 7.47 ± 0.79 -> 4.82 ± 0.22Improved

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