this post was submitted on 08 Oct 2023
507 points (97.0% liked)
Technology
58141 readers
4769 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
It is not "a flaw", it is the way language learning models work. They try to replicate how humans write by guessing based on a language model. It has no knowledge of what is a fact or not, and that is why using LLMs to do research or use them as a search engine is both stupid and dangerous
How would it hallucinate information from an article you gave it. I haven't seen it make up information by summarizing text yet. I have seen it happen when I ask it random questions
It does not hallucinate, it guesses based on the model to make you think the text could be written by a human. Personal experience when I ask into summarize a text. It has errors in it, and sometimes it adds stuff to it. Same if you for instance ask it to make an alphabetic a list of X numbers of items. It may add random items.
I've had it make up things if I ask it for a list of say 5 things but there's only 4 things worth listing. I haven't seen it stray from summarizing something I've fed it though. If its giving text, its been pretty accurate. Only gets funky when you ask it things where information isn't available. Then it goes with what you probably want