About Eigenvue
Eigenvue is a free, open-source interactive visualization platform for algorithms, deep learning, generative AI, and quantum computing. It provides step-by-step animated dissections with synchronized code highlighting and plain-language explanations — available in the browser, via Python, and via Node.js.
Why Eigenvue Exists
Understanding algorithms requires building spatial and temporal intuition that static diagrams and textbook pseudocode cannot provide. A bubble sort comparison, the propagation of gradients through a neural network, the rotation of a qubit on a Bloch sphere — these are inherently dynamic processes.
Eigenvue provides a unified visual language across four domains — classical computer science, deep learning, generative AI, and quantum computing — with dual distribution: a web application for instant browser access, and programmatic packages (pip and npm) for integration into notebooks, scripts, and educational toolchains.
22+ Visualizations Across Four Domains
From foundational sorting algorithms to quantum teleportation — each visualization is built with the same architecture and level of detail.
Classical Algorithms
7 algorithmsBinary Search, QuickSort, Merge Sort, BFS, DFS, Dijkstra's Algorithm, Bubble Sort
Deep Learning
5 algorithmsPerceptron, Feedforward Network, Backpropagation, 2D Convolution, Gradient Descent
Generative AI
5 algorithmsTokenization (BPE), Token Embeddings, Self-Attention, Multi-Head Attention, Transformer Block
Quantum Computing
5 algorithmsQubit Bloch Sphere, Quantum Gates, Superposition & Measurement, Grover's Search, Quantum Teleportation
Key Capabilities
Step-by-Step Playback
Play, pause, step forward, step backward, and scrub through every algorithm state at your own pace.
Synchronized Code Highlighting
Each step highlights the exact line being executed in pseudocode, Python, or JavaScript.
Plain-Language Explanations
Every step includes a human-readable explanation of what just happened and why it matters.
Custom Inputs
Supply your own arrays, graphs, or parameters and watch the algorithm adapt in real time.
Shareable URLs
Every visualization state is URL-encoded. Share the exact step you are looking at with a link.
Jupyter Notebook Integration
Embed interactive visualizations directly inside Jupyter notebooks with the Python package.
Install via pip or npm
Use the same visualizations programmatically in Python scripts, Jupyter notebooks, or Node.js applications.
Python Package
Python 3.10+
pip install eigenvue- Minimal dependencies — only Flask
- Optional Jupyter integration via IPython
- Serves interactive visualizations locally
Node.js Package
Node 18+
npm install eigenvue- Zero runtime dependencies
- TypeScript types included
- CLI binary available globally
Open Source on GitHub
Eigenvue is fully open source under the MIT License. Contributions, issues, and discussions are welcome.
Three-Layer Architecture
A clean separation between algorithm logic, data format, and rendering.
Generator Layer
TypeScript and Python functions that produce Step sequences for each algorithm.
Step Format
A universal JSON contract (v1.0.0) that serves as the interface between generators and renderers.
Rendering Layer
A Canvas 2D engine with layouts, animation, and playback controls that consumes step data.
Technology Stack
| Layer | Technologies |
|---|---|
| Web Application | Next.js 15, React 19, Canvas 2D, Tailwind CSS 4 |
| Python Package | Python 3.10+, Flask, Hatchling |
| Node Package | TypeScript, tsup, zero runtime dependencies |
| Testing | Vitest, pytest, JSON Schema validation |
| CI / CD | GitHub Actions — lint, typecheck, test, build, deploy |
| Documentation | Astro + Starlight |
License & Contributing
MIT License
Eigenvue is released under the MIT License. You are free to use, modify, and distribute it for any purpose — personal, educational, or commercial — with attribution.
Contributions Welcome
Bug reports, feature requests, new algorithm implementations, and documentation improvements are all welcome. See the Contributing Guide for development setup, branching strategy, and code standards.
Start Learning Today
No account required. No installation needed. Pick an algorithm and see it in action.