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

    Neuronpedia

    INDEX
    Explanations

    words related to things that are extremely difficult or impossible

    oai_token-act-pair · gpt-3.5-turbo

    phrases that express the concept of impossibility

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    oola
    -0.80
    notations
    -0.76
    lique
    -0.71
    rongh
    -0.68
    ingle
    -0.68
    endar
    -0.67
    anded
    -0.67
    holm
    -0.67
    along
    -0.65
    lance
    -0.65
    POSITIVE LOGITS
     feat
    0.81
     feas
    0.79
     conceivable
    0.73
     anymore
    0.72
     Impossible
    0.71
     comprehend
    0.70
     feats
    0.69
     impossible
    0.69
     unless
    0.69
     foreseeable
    0.69
    Act Density 0.044%

    No Known Activations