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

    Neuronpedia

    INDEX
    Explanations

    instances where someone is reaching out to others for various reasons

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     Tsukuyomi
    -0.93
    士
    -0.85
     Breakfast
    -0.76
     Attend
    -0.73
     Carbuncle
    -0.69
     diesel
    -0.69
    cake
    -0.68
     Bills
    -0.68
     Boxing
    -0.66
     FIGHT
    -0.65
    POSITIVE LOGITS
    stretched
    1.15
    haar
    0.91
    pring
    0.86
    worm
    0.85
    intent
    0.85
     solicit
    0.85
    bolt
    0.84
    seek
    0.84
    tips
    0.83
    strings
    0.82
    Act Density 15.704%

    No Known Activations