AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Stochastic Adaptive Activation Function

Kyungsu Lee, Jaeseung Yang, Haeyun Lee, Jae Youn Hwang

NeurIPS 2022 · Architectures and efficiency › RNNs and normalization · approach: Mechanistic circuit interpretability · 0 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
Replace ReLU with ASHmodel 1 · ImageNet 2012Top-1 accuracy (%): 76.4 -> 78.5Improved
Replace ReLU with ASHmodel 2 · ImageNet 2012Top-1 accuracy (%): 75.6 -> 78.6Improved
Replace ReLU with ASHmodel 3 · ImageNet 2012Top-5 accuracy (%): 90.7 -> 94.1Improved
Replace Swish with ASHResNet · CIFAR-10accuracy (%): 95.5 -> 95.7Improved
Replace ReLU with ASHWRN · CIFAR-10accuracy (%): 95.6 -> 96.7Improved
Replace ReLU with ASHDenseNet · CIFAR-10accuracy (%): 95.7 -> 96.0Improved
Replace ReLU with ASHResNet · CIFAR-100accuracy (%): 74.5 -> 76.5Improved
Replace Swish with ASHWRN · CIFAR-100accuracy (%): 78.9 -> 79.2Improved
Replace Softplus with ASHDenseNet · CIFAR-100accuracy (%): 84.1 -> 84.6Improved
Replace Swish with ASHYOLOv5 · COCOmAP@50: 74.0 -> 74.8Improved
Replace Swish with ASHYOLOv5 · PASCAL VOC detectionmAP@50: 72.6 -> 74.1Improved
Replace Swish with ASHU-Net · ADE20KmIoU (%): 51.3 -> 53.4Improved

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