News for hackers. Stuff that matters.

A comment on PolyThink: A Multi-Agent AI System to Eliminate Hallucinations. Parent comment

Re:PolyThink: A Multi-Agent AI System to Eliminate Hallucina... by TheKelsbee on Thursday April 17, 2025 @01:10PM (#43719639)
I have this same thought, and have tried similar approaches.

OP: Have you trained or fine tuned a model that specifically reasons the worker model inputs against the user input? Or is this basically just taking a model and turning the temperature down to near 0?

[ Parent | on HN ]
Re:PolyThink: A Multi-Agent AI System to Eliminate Hallucina... by kuberwastaken on Friday April 18, 2025 @08:58AM (#43727645)
Low temperature, heavy prompting to answer in a structured way. Sadly can't fine train models since this is API based but the approach does work!
[ Parent | on HN ]