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

    Neuronpedia

    INDEX
    Explanations

    infinitive verbs followed by certain words, like "be", "give", or "do"

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
    soDeliveryDate
    -0.84
    ////////////////////////////////
    -0.77
    calling
    -0.74
     Prediction
    -0.73
    quickShipAvailable
    -0.70
    usage
    -0.70
    cases
    -0.69
    ————————
    -0.67
     Lots
    -0.67
    lav
    -0.66
    POSITIVE LOGITS
     cooperate
    1.37
     accept
    1.36
     acknowledge
    1.35
     comply
    1.28
     obey
    1.26
     endorse
    1.26
     abide
    1.25
     bud
    1.24
     submit
    1.21
     disclose
    1.21
    Act Density 0.063%

    No Known Activations