News for hackers. Stuff that matters.
Re:Any insider takes on Yann LeCun's push against current ar... by josh-sematic on Friday March 14, 2025 @04:06PM (#43366675)
I don’t buy Lecun’s argument. Once you get good RL going (as we are now seeing with reasoning models) you can give the model a reward function that rewards a correct answer most highly, an “I’m sorry but I don’t know” less highly than that, a wrong answer penalized, a confidently wrong answer more severely penalized. As the RL learns to maximize rewards I would think it would find the strategy of saying it doesn’t know in cases where it can’t find an answer it deems to have a high probability of correctness.
Re:Any insider takes on Yann LeCun's push against current ar... by Tryk on Friday March 14, 2025 @04:13PM (#43366765)
How do you define the "correct" answer?
Re:Any insider takes on Yann LeCun's push against current ar... by jpadkins on Friday March 14, 2025 @04:22PM (#43366875)
obviously the truth is what is the most popular. /s
Re:Any insider takes on Yann LeCun's push against current ar... by josh-sematic on Friday March 14, 2025 @08:19PM (#43368708)
Certainly not possible in all domains but equally certainly possible in some. There’s not much controversy about the height of the Eiffel Tower or how to concatenate two numpy arrays.