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