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

    Neuronpedia

    INDEX
    Explanations

    numerals and words related to counting or measuring quantities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     OMITBAD
    -0.55
     two
    -0.52
    basicConfig
    -0.51
    two
    -0.49
     кӀ
    -0.48
    Three
    -0.47
     Two
    -0.46
    findOneAndUpdate
    -0.46
    gląda
    -0.45
     three
    -0.45
    POSITIVE LOGITS
     dozen
    1.73
     dozens
    1.36
    dozen
    1.35
     dizaines
    1.03
     dizaine
    0.91
    надцать
    0.87
     hundred
    0.87
     decenas
    0.87
    几十
    0.85
     HUND
    0.85
    Act Density 0.723%

    No Known Activations