Research map / Optimization / Optimizers
AdaBelief Optimizer: Adapting Stepsizes by the Belief in Observed Gradients
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| AdaBelief instead of Adam | ResNet18 · ImageNet val | top-1 accuracy: 63.79 -> 70.08 | Helped |
| AdaBelief instead of SGD | ResNet18 · ImageNet val | top-1 accuracy: 70.23 -> 70.08 | No clear effect |
| AdaBelief instead of Adam | SN-GAN · CIFAR-10 SN-GAN | FID: 13.05 -> 12.52 | Helped |
| AdaBelief instead of Adam | WGAN-GP · CIFAR-10 WGAN-GP | FID: 73.5 -> 61.8 | Helped |
| reduce Padam exponent p (less adaptivity) | Padam · CIFAR-10 WGAN | FID | Hurt |
| AdaBelief instead of Adam | Transformer · IWSLT14 | BLEU: 35.60 -> 35.85 | Helped |
| AdaBelief instead of AdaHessian | Transformer · IWSLT14 | BLEU: 35.90 -> 35.85 | No clear effect |
| AdaBelief instead of SGD | Faster-RCNN · PASCAL VOC | mAP: 79.50 -> 81.02 | Helped |
| AdaBelief instead of other optimizers | 2/3-layer LSTM · Penn Treebank | perplexity | Helped |
| AdaBelief instead of other optimizers | 1-layer LSTM · Penn Treebank | perplexity | No clear effect |
Closest papers by meaning
- A Comprehensive Framework for Analyzing the Convergence of Adam: Bridging the Gap with SGD (ICML 2025 · similarity 0.87)
- Non-asymptotic Analysis of Biased Adaptive Stochastic Approximation (NeurIPS 2024 · similarity 0.87)
- AdaFisher: Adaptive Second Order Optimization via Fisher Information (ICLR 2025 · similarity 0.86)
- Prodigy: An Expeditiously Adaptive Parameter-Free Learner (ICML 2024 · similarity 0.86)
- Adaptive Methods through the Lens of SDEs: Theoretical Insights on the Role of Noise (ICLR 2025 · similarity 0.86)
- AdaGrad under Anisotropic Smoothness (ICLR 2025 · similarity 0.86)
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.