OpenAI researchers argue that language models hallucinate because standard training and evaluation procedures reward guessing over admitting uncertainty (OpenAI)

https://ift.tt/w6fIuZK

OpenAI:
OpenAI researchers argue that language models hallucinate because standard training and evaluation procedures reward guessing over admitting uncertainty  —  Read the paper(opens in a new window)  —  At OpenAI, we're working hard to make AI systems more useful and reliable.

from Techmeme https://ift.tt/iIrAo3U