Tim Lee: Pretraining is imitation learning, post-training is reinforcement learning
binarybits · x · 2026-11-03
Tech writer Tim Lee (@binarybits), in a discussion with @batterystaple42, offers a crisp framing: pretraining is imitation learning, while post-training is (mostly) reinforcement learning. This echoes his broader argument that many still overestimate how much of a model's capabilities come from pretraining.
Related event: Tim Lee: Post-Training, Not Pre-Training, Drives LLM Capability(3 posts)→
More from AGI Musings
- 'Text-to-neural-network' vision: describe a task and get a purpose-built tiny net — flowersslop · 2026-11-03
- Micro schools will adopt AI tutors first and deliver '10x better' education, advocates say — RachelVT42 · 2026-11-03
- Yoav Goldberg: the 'of course AI can do that' bucket is getting insane — yoavgo · 2026-11-03
- HBR: The "Human-in-the-Loop" fallacy — why human review roles keep shrinking — luisdans · 2026-11-03
- Hot take: most AI slop is human slop with better grammar — AI is a multiplier — therealdanvega · 2026-11-03
- Naruto creator wrote 'in public' inside the manga, and that's what AI can't replace — marclou · 2026-11-03