Feynman Tutor vs How LLMs Work

Feynman Tutor is an AI skill (compatible with Claude Code, Cursor, and Windsurf) that inverts the typical AI tutoring model. Instead of the AI explaining concepts to you, you explain concepts to the AI — and the AI plays the role of a curious student, asking clarifying questions designed to expose the exact places where your understanding breaks down. It's the Feynman Technique implemented as an AI interaction pattern. The Feynman Technique — named after physicist Richard Feynman — is one of the most effective known learning methods: to understand something deeply, try to explain it simply enough that a child could understand. Where your explanation gets vague, evasive, or circular is exactly where the gaps are. Feynman Tutor automates the "curious student" role, generating targeted follow-up questions calibrated to probe the weak points in your explanation. The skill works by analyzing your explanations for hedging language, unexplained assumptions, circular definitions, and jumps in logic — then generating Socratic questions in response. It's designed to be used alongside active learning (reading a paper, working through a codebase) rather than as a standalone teacher. With 6 stars and created April 14, it's brand new — but it's a genuinely clever use of AI that prioritizes your understanding over AI-generated content.

"How LLMs Work" is a free, browser-based interactive guide that walks through the complete pipeline for building large language models — from raw web scraping to RLHF-trained conversational assistant. Created by Yash Narwal and based on Andrej Karpathy's technical deep-dive lecture, it's been getting significant traction on Hacker News (214+ points) for turning dense ML theory into something genuinely accessible. The site covers data collection and deduplication, Byte Pair Encoding tokenization with a live demo, pre-training and next-token prediction, inference with a probability sampling simulator, post-training with RLHF, and RAG. Each section uses animated visualizations, clickable pipeline diagrams, and canvas-based graphics — not static explainer images. The progressive narrative structure follows a single piece of text through every stage of the pipeline, making abstract concepts concrete. In an era where everyone uses LLMs but few understand how they work, this kind of high-quality educational resource matters for a different reason than tools: it raises the baseline competency of the entire developer ecosystem. Better-informed builders ask better questions, make better design decisions, and push vendors toward more transparency. This is the kind of project the HN community rewards — and deserves the signal boost.