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

    Neuronpedia

    INDEX
    Explanations

    specific phrases indicating progression in a text

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate conditional or hypothetical scenarios

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    PF
    -0.67
     MPG
    -0.63
    ãĥĨ
    -0.63
    ãĥ¼ãĥĨ
    -0.60
    ASED
    -0.59
    ectomy
    -0.58
    âĢ¢âĢ¢âĢ¢âĢ¢
    -0.58
    Interstitial
    -0.57
     Primary
    -0.57
     Marble
    -0.57
    POSITIVE LOGITS
    ratulations
    0.77
    aucus
    0.75
    oiler
    0.72
    ften
    0.69
     congr
    0.69
    endor
    0.68
    deck
    0.68
     parting
    0.67
    yah
    0.67
    hew
    0.66
    Act Density 0.145%

    No Known Activations