Research map / Architectures and efficiency / RNNs and normalization
Fractional Adaptive Linear Units
Julio Zamora, Anthony D. Rhodes, Lama Nachman
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Replace ReLU with FALU | Compact CNN · MNIST test | Top1%: 99.2 -> 99.3 | Improved |
| Replace ReLU with FALU | Compact CNN with pooling · MNIST test | Top1%: 99.0 -> 99.2 | Improved |
| Replace ReLU with FALU | ResNet18a · CIFAR-10 | Error%: 8.75 -> 7.36 | Improved |
| Replace GELU with FALU | ResNet18a · CIFAR-10 | Acc.%: 92.56 -> 92.64 | Improved |
| Replace ReLU with FALU | Compact CNN · Fashion-MNIST test | Top1%: 90.99 -> 91.72 | Improved |
| Replace Swish with FALU | ResNet18 · Fashion-MNIST test | Top1%: 96.00 -> 96.28 | Improved |
| Replace ReLU with FALU | VGG · Fashion-MNIST test | Top1%: 91.14 -> 92.09 | Improved |
| Replace ReLU with FALU | ResNet50 · ImageNet | Top1%: 75.8 -> 76.7 | Improved |
| Replace ReLU with FALU | ResNet50 · ImageNet | Top5%: 92.9 -> 93.28 | Improved |
Closest papers by meaning
- ADMM for Structured Fractional Minimization (ICLR 2025 · similarity 0.83)
- Unleashing the Potential of Fractional Calculus in Graph Neural Networks with FROND (ICLR 2024 · similarity 0.81)
- Minimum Width for Universal Approximation using Squashable Activation Functions (ICML 2025 · similarity 0.81)
- F-Adapter: Frequency-Adaptive Parameter-Efficient Fine-Tuning in Scientific Machine Learning (NeurIPS 2025 · similarity 0.81)
- FALQON: Accelerating LoRA Fine-tuning with Low-Bit Floating-Point Arithmetic (NeurIPS 2025 · similarity 0.81)
- Efficient Training of Neural Fractional-Order Differential Equation via Adjoint Backpropagation (AAAI 2025 · similarity 0.81)
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.
