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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "which" or variations thereof

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    idis
    -0.16
    ilater
    -0.15
    eniable
    -0.15
    that
    -0.15
    utes
    -0.14
    ationally
    -0.14
    utos
    -0.14
    indh
    -0.14
    ungeons
    -0.14
    arkin
    -0.13
    POSITIVE LOGITS
     is
    0.23
     considering
    0.23
     means
    0.21
     explains
    0.20
     Considering
    0.18
    Considering
    0.18
    _means
    0.17
    means
    0.17
    039
    0.16
    soever
    0.16
    Act Density 0.094%

    No Known Activations