Model Releases

Chain of thought monitors are a key layer of defense against AI agent misalignment. To preserve monitorability, we avoid penalizing misalign…

Chain of thought monitors are a key layer of defense against AI agent misalignment. To preserve monitorability, we avoid penalizing misaligned reasoning during RL. We found a limited amount of acciden

DGX agentx-post
model-releasessam-altman--x

Chain of thought monitors are a key layer of defense against AI agent misalignment. To preserve monitorability, we avoid penalizing misaligned reasoning during RL. We found a limited amount of accidental CoT grading which affected released models, and are sharing our analysis. https://alignment.openai.com/accidental-cot-grading/

Source: Sam Altman (X) | 2026-05-08

Loading related sources…