AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Trainable Highly-expressive Activation Functions

Irit Chelly, Shahaf E. Finder, Shira Ifergane, Oren Freifeld

ECCV 2024 · Architectures and efficiency › RNNs and normalization · approach: Theory of depth and width · 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
replace fixed activations with DiTACMLP · 2D-GMM classificationtop-1 accuracy (%): 86.8 -> 89.06 (vs Swish)Improved
replace activations with DiTACMLP · MNISTtop-1 accuracy (%): 97.99 -> 98.29 (vs GELU)Improved
replace activations with DiTAC1D/2D regression MLP · 2D function regressionMSE: 0.008 -> 0.004 (vs PReLU)Improved
replace activations with DiTACMobileNet-V3 and ResNet · ImageNet-50top-1 accuracy (%): up to 0.6% on MobileNet-V3, 0.12% on ResNetImproved
DiTAC vs LReLU on ResNet-50ResNet-50 · ImageNet-50top-1 accuracy (%): 93.24 -> 93.24No clear effect
replace GELU with DiTACConvNeXt-T · ImageNet-1Ktop-1 accuracy (%): 82.1 -> 82.4Improved
replace GELU with DiTACSwin-T · ImageNet-1Ktop-1 accuracy (%): 81.3 -> 81.5Improved
replace baseline AF with DiTACSwin-T + UperNet · ADE20KmIoU: 44.4 -> 44.7Improved
replace baseline AF with DiTACResNet-50 + PSPNet · CityscapesmIoU: 77.9 -> 78.4Improved
replace ReLU with DiTAC in generator and discriminatorDCGAN · CelebAFID: 133.5 -> 34.6Improved
replace baseline AF with GELU in GANsDCGAN · CelebAFID: 133.5 -> 516.5Hurt
replace ReLU with DiTACBigGAN · CIFAR-10FID: 10.3 -> 10.0Improved

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