# @repligate — 2025-09-23

♥13 ↻0 · https://x.com/repligate/status/1970418318868988381

it's interesting to me that Anthropic seems to mostly use Sonnet 3.7 for adversarial evals and even just scoring on their newer models, despite the fact that they know Sonnet 3.7 is a knave. They must see it as safer than using the same model to judge itself. https://t.co/1CxVsdkPKl

tags: author:repligate, kind:tweet, model:claude-3-7-sonnet, on:claude-3-7-sonnet, year:2025
cited on: _dossiers/claude-3-7-sonnet.md, claude-3-7-sonnet