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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "it" being mentioned, with a particular focus on cases where "it" is possibly involving an action or description

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
     Polk
    -0.69
     understatement
    -0.64
     Nights
    -0.62
     Siege
    -0.59
    IME
    -0.57
     Eyes
    -0.56
     Hanson
    -0.55
    Iowa
    -0.55
     Iowa
    -0.54
    do
    -0.54
    POSITIVE LOGITS
    alian
    1.23
     accordingly
    1.01
    unes
    0.97
    anwhile
    0.90
    self
    0.89
    atic
    0.88
     appropriately
    0.85
    atically
    0.82
     securely
    0.79
    ividual
    0.78
    Act Density 0.248%

    No Known Activations