Interactive Visualizations: How AI and LLMs work
Tokenization Explorer
See how AI breaks text into tokens in real-time. Understand why tokenization matters for API costs, context limits, and how different text types are processed by LLMs.
Intro to AI: Context Windows
Learn how LLMs manage and process information within limited memory windows. Understand the constraints and strategies that AI models use to handle conversations and long documents.
Vector Cloud Visualization
Discover how LLMs use semantic similarity and embeddings to understand context. See how adding specific details narrows the AI's focus to relevant information in 3D space.
The CRIT Method
Transform bad prompts into great ones using Context, Role, Interview, and Task. Watch how each element dramatically improves AI responses through interactive visualization.
Probability Distribution
Explore how LLMs calculate and choose between different possible next tokens. Visualize the probability layers that determine what the AI says next.
Token Generation Process
Follow the step-by-step journey of how LLMs generate text token by token. Understand the mechanics behind AI's text generation from start to finish.
Prompt Chaining
Learn how to chain prompts together to create powerful AI workflows. Take the output of one prompt and use it as input for the next, building loops and complex multi-step processes.
AI Hallucinations
Understand why AI confidently generates false information and how to detect it. Explore the knowledge gradient from public facts to company-specific data where hallucinations occur.
RAG Overview
Learn how Retrieval-Augmented Generation connects AI to external knowledge bases. See how RAG systems retrieve relevant information before generating responses to prevent hallucinations.
Data Refinery & Neural Network Training
Journey through the transformation of raw internet data into premium AI training material. Watch as billions of web pages are filtered, cleaned, and refined into high-quality datasets, then used to train neural networks.
Transformer Explainer
Dive deep into the architecture of transformer models with an interactive visualization. See how attention mechanisms, embeddings, and neural networks work together to power modern LLMs.