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

    Neuronpedia

    INDEX
    Explanations

    phrases suggesting recommendations or suggestions to try out something new

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    photo
    -1.21
    wordpress
    -1.00
    chat
    -0.97
    orge
    -0.97
    inburgh
    -0.94
    ixel
    -0.93
    quin
    -0.93
    operated
    -0.91
    vor
    -0.90
    apeshifter
    -0.89
    POSITIVE LOGITS
    itia
    1.04
    ?]
    1.00
     succumb
    0.99
    amaz
    0.98
     defer
    0.95
     indulge
    0.91
     acquies
    0.91
     concede
    0.89
     reinvent
    0.89
    ucc
    0.89
    Act Density 0.308%

    No Known Activations