How LLMs Work vs Ithihasas

"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.

Ithihasas (Sanskrit for "thus it was") is a web app for exploring characters, relationships, and narrative arcs across the Ramayana and Mahabharata. Built in a few hours as a Show HN project, it lets you browse the cast of these 100,000-plus-verse epics, understand how characters are connected, and follow story threads without reading the full texts. The app uses an AI layer to surface contextual information—relationships between characters, their roles in key episodes, family trees—in a digestible format. It's aimed at people who grew up with these stories culturally but find the full texts overwhelming, as well as researchers and curious outsiders wanting entry points. The project is a solo indie build with no monetization yet. At 126 HN points on launch day, it found a real audience. The comments included Sanskrit scholars praising the character mapping, parents looking for ways to share the stories with children, and diaspora users noting the gap it fills between formal academic resources and casual pop-culture summaries. Small project, real need.