Unrelated Response To A Question

#120
by sananaamir - opened

I asked HuggingChat if it could tell me about the underlying model that powers it and its training dataset. The model responded with a link to OpenAI research. Screenshot of conversation attached below.

image.png

Asking for clarification seemed to have not worked either haha.

image.png

Hugging Chat org

I asked HuggingChat if it could tell me about the underlying model that powers it and its training dataset. The model responded with a link to OpenAI research. Screenshot of conversation attached below.

Yes, that's a known bias of open models, they are hallucinating their creators as they cannot know.

sananaamir changed discussion status to closed

Sign up or log in to comment