News for hackers. Stuff that matters.
Re:QwQ: Alibaba's O1 Like Reasoning LLM by apocadam on Thursday November 28, 2024 @08:44PM (#42270082)
Exactly, it's completely arbitrary. I like to frame it in terms of fitting n points (the existing data in the problem) to a polynomial of degree n+1, where there's an infinite number of ways to pick a solution and still satisfy the initial data.
Re:QwQ: Alibaba's O1 Like Reasoning LLM by kaba0 on Friday November 29, 2024 @02:27PM (#42276363)
Maybe the "solution with the lowest Kolmogorov complexity".

In a sibling comment, I replied that usually a repeating pattern can also be applied, but that one usually requires storing the n-sequence, rarely making it the shortest encodable rule.