from Hacker News

  • Top
  • New

LRM reasoning breaks down down past an unfamiliarity threshold, not "complexity"

by k1m on 6/14/25, 10:54 PM with 0 comments