News for hackers. Stuff that matters.
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by pinoy420 on Saturday March 01, 2025 @02:45PM (#43222908)
A good prompt. You don’t just ask it. You tell it how to behave and give it a shot load of context
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by andix on Saturday March 01, 2025 @02:57PM (#43223020)
With Claude the context window is quite small. But with adding too much context it often seems to get worse. If the context is not carefully narrowly picked and too unrelated, the LLMs often start to do unrelated things to what you've asked.

At some point it's not really worth anymore creating the perfect prompt, just code it yourself. Also saves the time to carefully review the AI generated code.

Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by johnisgood on Saturday March 01, 2025 @04:37PM (#43223940)
Claude's context window is not small, is it not larger than ChatGPT's?
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by andix on Saturday March 01, 2025 @04:50PM (#43224071)
I just looked it up, it seems to be the rate limit that's actually kicking in for me.
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by johnisgood on Saturday March 01, 2025 @05:02PM (#43224170)
Yes, that's it! It is frustrating to me, too. You have to start a new chat with all relevant data, and a detailed summary of the progress/status.
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by johnisgood on Tuesday March 04, 2025 @01:57AM (#43251225)
(Because you reach the limit faster otherwise).
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by troupo on Saturday March 01, 2025 @02:58PM (#43223022)
Doesn't prevent it from hallucinating, only reduces hallucinations by a single digit percentage
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by copperroof on Saturday March 01, 2025 @03:22PM (#43223270)
Personally I’ve been finding that the more context I provide the more it hallucinates.
Re:Making o1, o3, and Sonnet 3.7 Hallucinate for Everyone by Rury on Saturday March 01, 2025 @04:11PM (#43223717)
There's probably a sweet spot. Same with people. Too much context (especially unnecessary context) can be confusing/distracting, as well as being too vague (as it leaves room for multiple interpretations). But generally, I find the more refined and explicit you are, the better.