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

    Neuronpedia

    INDEX
    Explanations

    action-related words, specifically those related to pursuing or following someone or something

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    ocker
    -0.99
    icka
    -0.94
    activation
    -0.90
    ittered
    -0.89
    optional
    -0.87
    aston
    -0.86
    ifier
    -0.86
    isphere
    -0.86
    ifications
    -0.85
    amination
    -0.85
    POSITIVE LOGITS
    eers
    1.11
     pursued
    1.06
    crow
    1.04
     pursue
    1.00
    ttes
    0.98
    ries
    0.94
    rily
    0.94
    hunt
    0.94
     HUN
    0.93
    suit
    0.91
    Act Density 1.069%

    No Known Activations