News for hackers. Stuff that matters.

A comment on Fixing Hallucination with Knowledge Bases. Parent comment

Re:Fixing Hallucination with Knowledge Bases by ajcp on Thursday May 04, 2023 @08:30PM (#35823924)
Guessing.

When we give information we don't know to be true -which is what the model does when it builds low confidence tokens on each other- we call it guessing. Why would this be any different?

[ Parent | on HN ]
Re:Fixing Hallucination with Knowledge Bases by eximius on Thursday May 04, 2023 @08:35PM (#35823972)
Guessing has a very mild implication of thought or estimation from a model of understanding of the subject in question.

Hallucination, on the other hand, is evocative of a less reasonable/rational source.

[ Parent | on HN ]
Re:Fixing Hallucination with Knowledge Bases by omneity on Thursday May 04, 2023 @09:39PM (#35824333)
That would actually be an educated guess or a guesstimate. A guess is as good as any.
[ Parent | on HN ]
Re:Fixing Hallucination with Knowledge Bases by eximius on Friday May 05, 2023 @12:15AM (#35825157)
Maybe strictly definitionally, but comparing the connotations of the two, I assume people don't actually truly randomly guess.
[ Parent | on HN ]
Re:Fixing Hallucination with Knowledge Bases by duckmysick on Friday May 05, 2023 @04:08AM (#35826560)
Isn't that anthropomorphic too? It's humans doing guessing, not objects.
[ Parent | on HN ]