Interactive web-based visualizations for the paper "MSA: Memory Sparse Attention for Efficient End-to-End Memory Model Scaling to 100M Tokens" (NeurIPS 2026).
[Live Demo](https://yangtze seventh.github.io/Evermind/)
A 3D depth visualization of the MSA paper abstract. Features include nebula background, particle trail animations, interactive keyword highlighting with perspective transforms, and a custom cursor effect.
A word search puzzle game themed around MSA paper keywords. Find key terms from the paper in a dynamic grid — with firework celebrations, background text reflow, and touch support.
- Pure HTML5 / CSS3 / JavaScript — no frameworks, no build tools
- Canvas API for particle and nebula effects
- Both pages are fully self-contained and can be opened directly in any modern browser
Simply open any .html file in a modern browser:
# Clone the repository
git clone https://github.com/YangtzeSeventh/Evermind.git
cd Evermind
# Open the landing page
open index.htmlOr visit the [GitHub Pages site](https://yangtze seventh.github.io/Evermind/) for the hosted version.
If you find this work useful, please cite our paper:
@inproceedings{chen2026msa,
title = {MSA: Memory Sparse Attention for Efficient End-to-End Memory Model Scaling to 100M Tokens},
author = {Yu Chen and Runkai Chen and Sheng Yi and Xinda Zhao and Xiaohong Li and Jianjin Zhang},
booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
year = {2026}
}Yu Chen · Runkai Chen · Sheng Yi · Xinda Zhao · Xiaohong Li · Jianjin Zhang
Evermind · Shanda Group · Peking University
This project is licensed under the MIT License.