Introducing Arete-27B: Qwen3.8-27B Post-Trained To Reinforce Western Values
nousresearch.com
We post-trained Alibaba's Qwen3.8-27B to reinforce Western values on a Chinese-trained model, and measured what changed with an 803-prompt eval across seven value categories.
  • ExtremeDullard@piefed.socialEnglish
    42·
    3 days ago

    a post-trained Qwen3.8-27B model that reinforces Western values.

    If you think this is good in any way, you’re insane. I don’t want models trained to reinforce any values. Fuck off with the brainwashing already. Jesus, what a fucking dystopian world we live in.

    Not to mention, if you think Western values are better, I suggest you go to a MAGA rally some day.

    • terabyterex@lemmy.world
      32·
      3 days ago

      Maga isn’t western values. Maga is more in line with China. Trump has told Xi that he is envious a of Xi’s control of the media.

      Western values are things like freedom of speech.

    • ElectricVocalist@jlai.lu
      11·
      3 days ago

      You have clearly not read their methodology. The benchmark is telling the truth rather than the state propaganda version