303 - How LLMs Work - the 20 minute explainer
From Fragmented - AI Developer Podcast by Kaushik Gopal, Iury Souza
February 2, 2026 · 26 min · Episode 303
About this episode
The episode explains how LLMs work, covering the full pipeline from tokenization to inference.
Ever get asked "how do LLMs work?" at a party and freeze? We walk through the full pipeline: tokenization, embeddings, inference — so you understand it well enough to explain it. Walk away with a mental model that you can use for your next dinner party.
People in this episode
Hosts: Kaushik Gopal, Iury Souza
Topics covered
- LLMs
- tokenization
- embeddings
- inference
- mental model
- explanation
Keywords
- LLMs
- tokenization
- embeddings
- inference
- explanation
- mental model
Mentioned in this episode
Books & works: LLMs
More episodes of Fragmented - AI Developer Podcast
- 310 - Mitchell Hashimoto on Ghostty & His Agentic Coding Workflow · April 14, 2026 · 60 min
- 309 - Background Agents · April 1, 2026 · 26 min
- 308 - How Image Diffusion Models Work - the 20 minute explainer · March 24, 2026 · 25 min
- 307 - Harness Engineering - the hard part of AI coding · March 17, 2026 · 30 min
- 306 - Keeping your agent instructions in sync and effective · March 10, 2026 · 23 min
- 305 - Subagents explained - What they are, when (not) to spawn them · February 17, 2026 · 27 min
Explore listener stats, chart rankings, contacts and more on the Fragmented - AI Developer Podcast podcast page.