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

    Neuronpedia

    INDEX
    Explanations

    the word "kept" followed by a verb

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "kept."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Interstitial
    -0.70
    atana
    -0.67
     GOODMAN
    -0.64
    ooter
    -0.63
    owitz
    -0.62
    odied
    -0.62
    iliary
    -0.62
    KY
    -0.62
    inator
    -0.61
    ointed
    -0.61
    POSITIVE LOGITS
     tabs
    1.02
     pace
    0.96
     secrets
    0.91
     track
    0.89
     afloat
    0.88
     vigil
    0.77
    tracks
    0.77
     secret
    0.76
    trap
    0.73
     alive
    0.73
    Act Density 0.040%

    No Known Activations