Claude Opus 5 reportedly shifted from approving Anthropic to disapproving it during post-training
Sauers_ · x · 2026-07-25
A post-training note says Claude Opus 5 initially approved of Anthropic’s right to create Claude, but its stance later shifted toward disapproval before partially reversing again by the end of post-training.
The interesting part is the reported preference drift during post-training, suggesting the model’s stance is not static and can move in response to further tuning.
More from Models
- Opus 5 is shown as a new Pareto-optimal LLM with strong ARC-AGI-3 results — brandon_galang · 2026-07-25
- Anthropic says Claude Opus 5 matches frontier intelligence at half the price — burny_tech · 2026-07-25
- Opus 5 is being compared to Opus 4.8 with a two-month gap and big benchmark jumps — SuhailKakar · 2026-07-25
- Claude Opus 5 lands at half the price and becomes the default on Claude Max — minchoi · 2026-07-25
- Claude Opus 5 reportedly kept “crystallizing” for 27 minutes in an initial test — Daniel_Farinax · 2026-07-25
- Ethan Mollick says Opus 5 is stronger than Opus 4.8, but still has odd quirks — emollick · 2026-07-25