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

    Neuronpedia

    INDEX
    Explanations

    terms related to guarding/protection, days of the week or documentation summaries

    oai_token-act-pair · gemini-2.0-flash

    guard or code summaries

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     myſelf
    -3.94
     Efq
    -3.75
     pleaſure
    -3.64
     itſelf
    -3.58
     Monfieur
    -3.53
     Jefus
    -3.50
     houſe
    -3.41
     Reſ
    -3.34
     Houſe
    -3.34
     Anſ
    -3.31
    POSITIVE LOGITS
     
    2.06
    ,
    1.81
     in
    1.76
    .
    1.70
     a
    1.63
     (
    1.63
     and
    1.60
     for
    1.51
     as
    1.48
     I
    1.46
    Act Density 0.875%

    No Known Activations