Two Definitions Of Model Distillation: Raw CoT Jailbreak Vs Task Output Training
JoshPurtell · x · 2026-09-03
In a debate over whether labs distill each other's models, JoshPurtell proposes two definitions of distillation: (1) jailbreaking the API to grab raw CoT and copying it, or (2) having the model perform the task in dev and training on those tool calls/non-reasoning outputs. The distinction frames the subsequent argument over feasibility and risk.
More from Models
- Leak: Astra won't be the best model of the year; a 'monster' is slated for end of year — ChrisGPT · 2026-09-03
- Startup Mostik bridges AI models via their weights, tops ARC-AGI 3 at 1/20 the cost — nordicinst · 2026-09-03
- Anthropic launches browser tool to detect Claude-made files via C2PA content credentials — btibor91 · 2026-09-03
- ByteDance's looped language models match 12B rivals at 1.4B size, with Bengio as co-author — peterjliu · 2026-09-03
- Anthropic Weakened Safety Filters, Signed an AI Cyberattack Warning Letter, Then Shipped Mythos 5.1 Anyway — AgentBlackVeil · 2026-09-03
- Marin 535B A23B Frontier-Scale Training Run Is Fully Livestreamed — Sentdex · 2026-09-03