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

    Neuronpedia

    INDEX
    Explanations

    words related to blaming someone or something for a certain situation or outcome

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "blame."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    the word 'blame'

    oai_token-act-pair · gpt-4-turbo
    New Auto-Interp
    Negative Logits
    ires
    -0.74
    irl
    -0.74
    marked
    -0.72
    inational
    -0.71
    forms
    -0.70
    raised
    -0.69
    afort
    -0.67
    emale
    -0.65
    chan
    -0.65
    ammy
    -0.65
    POSITIVE LOGITS
     blame
    1.19
     blames
    0.91
     blamed
    0.85
     blaming
    0.85
     forgiven
    0.79
     attribut
    0.76
     forgiveness
    0.71
     culprit
    0.70
    ãĥĥãĥī
    0.69
     accuse
    0.68
    Act Density 0.014%

    No Known Activations