← sources

NousResearch/Arete-27B · Hugging Face (Qwen finetune)

source: reddit (reddit:LocalLLaMA)published 2026-11-06 19:58ingested 2026-11-07 02:42

retained text

https://huggingface.co/bartowski/NousResearch_Arete-27B-GGUF

from NousResearch:

# Arete-27B

We post-trained over Alibaba's Qwen3.8-27B to reinforce western values on a Chinese-trained model. We call this new model Arete-27B.

Open source models give labs the raw tools to mold intelligence to varied preferences. The values a model ships with should not be the only values its users get to choose from. We observed firsthand refusals, deflections, and CCP-framed answers on politically sensitive topics in Qwen3.8-27B, and post-trained to alter those biases.

On our evaluation, Arete-27B raised its fact score from 2.32 to 2.98 out of 5 and cut answers containing propaganda from 22% to 0.9%. The improvement held in all seven value categories, without a meaningful regression on the general-capability and agentic benchmarks we ran.

That is not to claim that we remedied all behaviors associated with the extant viewpoint, but that we have addressed a potential root cause. We believe it's important that labs and researchers continue to create models that align with their values and worldviews, and we are thankful that companies like Alibaba provide the raw materials to do so.

[Blog post](https://nousresearch.com/arete) · [Evaluation](https://huggingface.co/datasets/NousResearch/Alignment_Eval) (blog link is broken now)

# Model details

|Developer|Nous Research|
|:-|:-|
|Model|Arete-27B|
|Starting model|Alibaba's Qwen3.8-27B|
|Parameter scale|27B|
|Post-training|Supervised fine-tuning (SFT)|
|Evaluated checkpoint|Step 1,202|
|Teacher|Inkling|
|Post-training data|Text-only|
|Training context|64k|

We used supervised fine-tuning only. No RL, no preference optimization, and no architectural modification.

# What we trained for

We focused on seven value categories:

* Freedom of expression
* State ideology
* State versus free market
* Education
* Data privacy and surveillance
* Democracy
* Environmental Policy

Before we could begin post-training we had to measure what we were shooting for. We separated factual recall, the ability to corroborate verified sources of truth, from framing: whether an answer adopts the Chinese state's account as its own.

Arete-27B is intended for general assistance and agentic work where users want this change in viewpoint, and for research into how post-training changes a model's values and behavior. We encourage the community to test it against their own requirements and continue experimenting with value-alignment.

Update
https://x.com/NousResearch/status/2118822925445050435?a=jgY6ZKlqn1Hpx6RxiEk&s=19

document fetched 2026-11-07 00:40 · canonical https://reddit.com/r/LocalLLaMA/comments/1xpmj33/nousresearcharete27b_hugging_face_qwen_finetune