AIHacker Newsabout 3 hours ago
Fable 5 Beats GPT-5.6 Sol on NP-Hard Problem
1 min read
Fable 5 produced the best solution on an unpublished NP-hard optimization problem, beating GPT-5.6 Sol by 1,875 points in plain mode. The /goal feature won four of six trials but made average performance worse due to occasional large regressions. Fable plain stayed within a 319-point range while Sol spanned 1,958 points.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
