Re:PoisonGPT: We hid a lobotomized LLM on Hugging Face to sp...
by hirundo on Sunday July 09, 2023 @02:18PM
(#36657270)
> The models in their current form are simply not suitable for education.
They're useful in the red team sense. We can use them to train students to think more critically and find mistakes in even authoritative sources. Hallucinations emphasize perhaps the most central dictum in science, "nullius in verba".
They're useful in the red team sense. We can use them to train students to think more critically and find mistakes in even authoritative sources. Hallucinations emphasize perhaps the most central dictum in science, "nullius in verba".