News for hackers. Stuff that matters.
Re:Ask HN: Strategies to Reduce AI Hallucinations? by dheera on Monday July 29, 2024 @02:12AM (#41098274)
Explicitly allow it the option to be unsure, e.g. "If you do not know the answer, respond with 'none'" or "If you are unsure of the answer, just say that", etc.

Otherwise it does what humans do when asked interview questions, they bullshit because if you bullshit is a 20% chance of landing the job, whereas if you say "I don't know" there is a 0% chance of landing the job. The kind of RLHF training that was put into ChatGPT probably replicates a similar reward structure.

Re:Ask HN: Strategies to Reduce AI Hallucinations? by Havoc on Monday July 29, 2024 @02:59AM (#41098436)
This assumes the it “knows” what it knows ahead of time which isn’t true. It’s why models struggle to tell you how many Rs are in strawberry or why it can’t tell you how many words are in its response. Also doesn’t really have a concept of certainty beyond perhaps logprobs which are flimsy indicators at best and the model isn’t inherently aware of them
Re:Ask HN: Strategies to Reduce AI Hallucinations? by konschubert on Monday July 29, 2024 @03:12AM (#41098476)
I think you're overthinking this.

If you tell it that it can be unsure, the likelihood that an answer is correct increases. Make of that philosophically what you will. If it works it works.

Re:Ask HN: Strategies to Reduce AI Hallucinations? by dheera on Monday July 29, 2024 @02:56PM (#41102771)
"Prove Fermat's Last Theorem using game theory"

https://i.imgur.com/XbLanp1.png

"Prove Fermat's Last Theorem using game theory. If you think this is a bullshit question or are unsure, please just say that."

https://i.imgur.com/knbaPcq.png