Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights
natolambert · x · 2026-10-27
Responding to questions about Cohere's model, AI researcher Nathan Lambert offers his "simplest explanation": it's narrow RL training reliant on Cohere's internal infrastructure, which isn't designed well for an open-weight release. He jokes he's about to get five press requests about it — an insider read suggesting the model's capabilities are coupled to training infra that doesn't translate to open weights.
More from Models
- Leaked Terminal-Bench 4.0 chart shows GPT-6.1 Terra leading at lowest cost, unverified — Angaisb_ · 2026-10-27
- OpenAI brings GPT-6.1 Terra to the API and Codex for fast, cheap computer use — OpenAIDevs · 2026-10-27
- GPT-6.1 Terra rolls out to all Plus, Pro, Enterprise and Edu users — OpenAI · 2026-10-27
- In cyber, intention matters little: Chintala on model info leaks, and what users should do — brhydon · 2026-10-27
- Why don't more LLMs ship 10M-token context? A researcher asks about attention tradeoffs — agarwl_ · 2026-10-27
- Ortet's first health eval: 34 open-weight models, 50 tasks, generic rankings fall short — thegautamkamath · 2026-10-27