What happened
Nous Research fine-tuned Alibaba’s Qwen3.8-27B on about 53 million tokens to create Arete-27B. The company reports that the model’s propaganda rate fell from 22 percent to 0.9 percent, while its factual accuracy score rose from 2.32 to 2.98 out of 5. On Hermes Bench, Arete-27B scored 65.60 percent versus 65.77 percent for the base model; its weights, test dataset, and training data are available under an Apache 2.0 license on Hugging Face.
Why it matters
The reported benchmark results compare propaganda rates before and after fine-tuning, while the unchanged Hermes Bench score offers a concrete measure of capability retention in this evaluation.