On-Policy Distillation: LLM's Post-Training Meta of 2026

bycloud · 44,253 views

Explainer on on-policy distillation, a key 2026 post-training technique for builders tuning smaller models.

More in Models
1:59Models
GPT-6.2 Astra and GPT-6.2 Sol—Models built for the real world
OpenAI · 333,228 views · Nov 2026
OpenAIGPT-6.2launch
19:38Models
2026 in LLMs — Simon Willison
AI Engineer · 30,656 views · Nov 2026
Simon WillisonAI Engineerkeynote
37:25Models
How to compress text to 1 bit per character | Compression is Intelligence Part 3
3Blue1Brown · 865,342 views · Nov 2026
compressionexplainertutorial