Aether AI Visualizations Lab
A professional-grade VS Code extension for AI/ML visualization, interactive code execution tracing, and structured programming education.
Aether AI Visualizations Lab reveals the inner workings of your code and machine learning models — transforming abstract algorithms into real-time, interactive diagrams. Whether you are debugging a recursive function, studying gradient descent, or learning how a Transformer attends to tokens, this extension makes the invisible visible.
Key Features
- 47 Interactive AI/ML Visualizations: Explore neural networks, transformers, GANs, and more directly inside VS Code.
- Complete Code Execution Tracing: See every variable, stack frame, and heap object at every line of execution.
- AI-Powered Explanations: Get natural-language insights on loops, memory, and performance powered by Google Gemini.
- Structured Learning: Built-in curriculum covering Python fundamentals through advanced DSA.
- Voice-First Interaction: Ask questions about your code using your voice and get instant AI answers.
Quick Start
- Install: Search for "Aether AI Visualizations Lab" in the VS Code Extensions panel.
- Visualize: Right-click in any supported file and choose "Visualize".
- Explore: Use the floating controls to step through execution or click the sidebar to explore AI/ML concepts.
AI/ML Interactive Visualizations
Aether AI Visualizations Lab includes 47 fully interactive concept visualizations across several categories:
- ML Fundamentals: Gradient Descent (2D/3D), Linear/Logistic Regression, Loss & Activation Functions.
- Deep Learning: Neural Networks, Backpropagation, CNNs, Object Detection.
- Transformers & NLP: Multi-head Attention, BERT, GPT, Word Embeddings, LSTMs.
- Generative Models: GANs, Diffusion Models.
- Reinforcement Learning: Q-Learning, Policy Gradients, RLHF.
- Classical ML: Decision Trees, SVM, K-Means, PCA.
- Agentic AI: ReAct Agents, RAG, Vector Databases, Multi-Agent Systems.
Code Execution Visualizer
Transform your code into a live diagram.
- Stack & Heap: Visual cards for function frames and heap objects with "Box & Pointer" reference arrows.
- Step Navigation: Playback controls, sliders, and auto-play to navigate your code's history.
- Special Modes:
- LENS: Magnify complex data structures.
- Turbo: High-speed tracing for large algorithms.
- Cumulative: Keep historical frames visible for recursion analysis.
- Live Sync: Automatically re-traces your code as you type (optional).
Learning Curriculum
The built-in course browser offers structured paths for:
- Python (24 Sections): From basic data types to advanced DSA patterns and real-world projects.
- Java, C++, & C#: Core programming fundamentals and language-specific features.
- Challenges: Mini-projects including CLI utilities, web scraping, and automation.
Configuration & Commands
Important Settings
aetherLab.cumulativeMode: Keep all historical stack frames visible.aetherLab.autoSyncEnabled: Automatically re-trace on every edit (off by default).aetherLab.maxTraceSteps: Limit the number of execution steps (default: 1000).
Main Commands
- Aether AI Lab: Visualize: Start tracing the active file.
- Aether AI Lab: Open: Open the home panel.
- Aether AI Lab: Ask AI: Send code to AI for insights.
Troubleshooting
- Indefinite Loading: Ensure the required language runtime (Python, JDK, etc.) is in your system PATH.
- High CPU Usage: Increase
aetherLab.debounceDelay or disable aetherLab.autoSyncEnabled.
- Instruction Limit: Increase
aetherLab.maxTraceSteps if your code has many steps.
For technical architecture and build instructions, see CONTRIBUTING.md.
Developed by Soubhik
| |
