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

    Neuronpedia

    INDEX
    Explanations

    phrases related to reasons or explanations

    oai_token-act-pair · gpt-3.5-turbo

    causal explanations or justifications in statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    achable
    -0.79
    ipers
    -0.71
    Deal
    -0.64
    iveness
    -0.63
    mbol
    -0.62
    venge
    -0.62
    scribe
    -0.62
    rouse
    -0.61
    urity
    -0.61
    holy
    -0.60
    POSITIVE LOGITS
     although
    0.94
     "[
    0.82
     despite
    0.79
     while
    0.79
     unlike
    0.78
     respondents
    0.77
     neither
    0.76
     whilst
    0.75
    ecause
    0.75
     Lack
    0.74
    Act Density 0.418%

    No Known Activations