zlacker

[parent] [thread] 1 comments
1. logicc+(OP)[view] [source] 2023-12-27 15:11:19
>Chinese LLMs are clearly trained to avoid answering certain topics that their government deems sensitive

But they're not; you can download open source Chinese base models like Yi and Deepseek and ask them about Tianmen Square yourself and see, they don't have any special filtering.

replies(1): >>meowfa+R1
2. meowfa+R1[view] [source] 2023-12-27 15:21:32
>>logicc+(OP)
I suspect they will crack down on that within the next few years.
[go to top]