Discuss: Nous Research strips state propaganda from Qwen model with light fine-tuning

This is the discussion thread for our news brief Nous Research strips state propaganda from Qwen model with light fine-tuning.

Nous Research says a small supervised fine-tune cut propaganda in Alibaba's Qwen3.8-27B to under one percent, a result relevant to teams building on open models.

What does this change for you or your team? Share your take.