Episode
"LLMs are (still) mostly powered by imitative learning, not RL" by Steven Byrnes
- Published
- Jul 26, 2026
- Duration seconds
- 1168
- Processing state
not_requested
Actions
POST https://stenobird.com/v1/public/podcasts/lesswrong-curated-popular-5643401/episodes/llms-are-still-mostly-powered-by-imitative-learning-not-rl-by-steven-byrnes/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/lesswrong-curated-popular-5643401/llms-are-still-mostly-powered-by-imitative-learning-not-rl-by-steven-byrnes.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
Reinforcement learning from verifiable rewards (RLVR) is the hot new thing in LLM training. It's so hot, and people spend so much time talking about it, that they sometimes lose sight of the big picture. Stepping back, LLMs can do lots of very impressive things. How? Where did those capabilities come from? Fundamentally, they come from a combination of: (1) Imitative learning, including pretraining and supervised fine-tuning (SFT) See my earlier discussion: “LLM pretraining magically tra...