Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    words that can be numbers

    oai_token-act-pair · gemini-2.0-flash

    numerical values and large quantities

    np_token-act-pair-logits · gpt-4o-mini

    magnitude numbers

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    MessageTagHelper
    -0.85
    ArrowToggle
    -0.66
     ApiException
    -0.66
    Naissance
    -0.60
     ExecuteAsync
    -0.57
    scriptcase
    -0.52
    formik
    -0.49
     الرياضيه
    -0.47
    istoitu
    -0.47
    eches
    -0.46
    POSITIVE LOGITS
     million
    1.30
     hundred
    1.22
     thousand
    1.16
     billion
    1.09
    hundred
    1.08
     dozen
    1.03
    thousand
    0.98
     Million
    0.98
    million
    0.96
     lakh
    0.94
    Act Density 3.898%

    No Known Activations