AILLM
Beam Search vs Sampling: How LLMs Decode
Compare greedy, beam search, top-p sampling, and contrastive search for LLM decoding. Interactive playground shows each strategy picking tokens live.
9 min read
Loading...
Back to all posts
Category
AI
Teach machines to think, learn, and surprise you
56 posts
Showing 13-24 of 56 posts
AILLM
GGUF vs GPTQ vs AWQ: Pick the Right Format
GGUF, GPTQ, and AWQ are three ways to shrink LLM weights. Each format makes different tradeoffs between hardware flexibility, accuracy, and speed.
9 min read
AILLM
Temperature & Top-p: How LLMs Choose Words
Learn how temperature and top-p control LLM output. Interactive playground lets you tune both and watch the probability distribution change.
9 min read
AILLM
How Transformers Work Step by Step
Understand how transformers process text from token to prediction. An interactive guide with a live playground to trace the full forward pass.
10 min read
AI
What Are Embeddings and Why Do They Matter
What are embeddings? See how AI turns text into numbers. Interactive visualizer compares phrases and reveals why similar meanings become nearby vectors.
7 min read
AI
Why You Need a Vector Database, Not a List
Why can't you just use a Python list for vector search? Interactive simulator shows how brute force dies at 100K vectors and why you need a vector DB.
7 min read
AI
HNSW: How Vector Search Actually Works
HNSW finds nearest neighbors in milliseconds by navigating a multi-layer graph. Interactive navigator shows the search powering every vector database.
8 min read
AILLM
How to Add Memory to Any LLM
LLMs forget everything between API calls. Buffer, summary, and vector memory fix this. Interactive simulator shows what each strategy remembers.
9 min read
AILLM
Context Engineering for Coding Agents
See what Copilot, Cursor, and Claude Code feed their LLMs. Interactive agent simulator lets you build context for a coding task and compare strategies.
9 min read
AILLM
LLM Function Calling with MCP: A Practical Guide
Learn how MCP tools turn Python type hints into JSON Schema. See how LLMs pick the right tool and build better definitions with FastMCP.
10 min read
AILLM
Build an AI Agent API with FastAPI
Deploy an AI agent as a REST API with FastAPI in Python. 11 steps: endpoints, streaming, chat memory, auth, Docker, and unit tests.
14 min read
AILLM
Build a ReAct AI Agent from Scratch in Python (Step-by-Step)
Build a ReAct (Reasoning + Acting) AI agent from scratch in pure Python. 10-step interactive tutorial with runnable code, output previews, and hands-on challenges.
10 min read