That's just conjecture though surely. Like not saying you're wrong but the same thing would result from crawling and using results from Chinese Web for training along with everything else. If the most prevalent model people are posting responses from online in Chinese is qwen then of course that would happen. There's no evidence to say either way really.
AI models. If you ask Claude Opus in Chinese which model it is, sometimes it says Qwen.
That's just conjecture though surely. Like not saying you're wrong but the same thing would result from crawling and using results from Chinese Web for training along with everything else. If the most prevalent model people are posting responses from online in Chinese is qwen then of course that would happen. There's no evidence to say either way really.
It's the exact same reason people conjecture that Qwen distills Claude.
All models are trained on outputs from all other frontier models at training time.
So you mean that they have no problem distilling but are crybabies and also accuse the chinese of doing it?
If it’s a question for me I’m not sure. I guess I’d like to ask what prompted this question?
I read 'we' (present tense) rather than 'we'll' (future tense). So, nevermind!