THE VIDEO SHELF · large language models
Deep Dive into LLMs like ChatGPT
Andrej Karpathy's three-hour-plus general-audience deep dive into how ChatGPT-style LLMs are built, covering the full training stack from pretraining to post-training. It also gives mental models for their quirks, like hallucination, and how to use them well.
Why HN readers recommended it
HN's most frequent recommendation for a clear, approachable foundation in how LLMs work without heavy math.
This is a summary from the source dataset, not an endorsement by Hacker News or the video creator.
Video & recommendation details
- Creator / channel
- Andrej Karpathy
- Topic
- large language models
- Format
- lecture
- Level
- intro
- HN recommenders in dataset
- 10
- HN mentions in dataset
- 18
- Recommendation years in dataset
- 2025–2026
These counts and years describe the included Hacker News records; they are not YouTube views or the video’s publication date.
From the original HN conversations
Excerpts from the recorded comments. Follow each source for the full comment and surrounding discussion.
I think you'll get a lot of use from this video from Andrej Karpathy: [VIDEO] It is long, but don't get scared off. He goes over a ton of different stuff related to model training, but makes it very easy to understand.
InvidFlower on Hacker News · 2025-03
In Ladder: Self-improving LLMs through recursive problem decompositionDepending on how much time you have, the first thing I would recommend is to watch Karpathy's 'deep dive into LLMs' video ( [VIDEO] ). It is 3.5 hours but it will give you so much instinctive understanding of how these things work and will …
GodelNumbering on Hacker News · 2025-04
In Show HN: I made a free tool that analyzes SEC filings and posts detailed reportsSince you're posting here, you're looking for the shortcut. The shortcut is Karpathy's "Let's Build GPT: from scratch, in code, spelled out" video: [LINK] Then there is a good video that dives into LLMs and how they work that is quite approachable: [VIDEO] From there, …
runjake on Hacker News · 2025-10
In Ask HN: Build Your Own LLM?
Summaries, classifications and recommendation records are from the HackerTube source dataset. Read about this library and its limitations.

