Interactive visualizations and paper analyses of optimization algorithms and deep learning concepts, built with Python (Matplotlib, Manim) and HTML/Three.js.
| Paper | Topic | HTML | Manim Code | Manim Video |
|---|---|---|---|---|
| Adam: A Method for Stochastic Optimization (Kingma & Ba, 2015) | Adaptive moment estimation optimization | adam.md | adam_manim.py | MP4 |
| Deep Residual Learning for Image Recognition (He et al., CVPR 2016) | Residual networks (ResNet) | resnet.md | resnet_manim.py | MP4 |
| Why does deep and cheap learning work so well? (Lin, Tegmark & Rolnick, 2017) | Physics of deep learning | cheap.md | cheap_manim.py | MP4 |
| Attention Is All You Need (Vaswani et al., NIPS 2017) | Transformer architecture | attention.md | attention_tofix_01.py | MP4 |
- Adam Slow Motion Comparison - Step-by-step Adam optimization visualization
- Adam Integration - Interactive Adam parameter integration
- Full paper analyses with mathematical derivations (TeX/MathJax)
- Python code with highlighting and line numbers
- Manim animation videos for key concepts
- Interactive HTML demos
This repository automatically deploys to GitHub Pages via GitHub Actions on every push to main.