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

    Neuronpedia

    INDEX
    Explanations

    phrases related to analysis, critique, and evaluation

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    oulos
    -0.84
    retty
    -0.79
    stories
    -0.74
    abee
    -0.74
    itary
    -0.72
    gate
    -0.72
    emale
    -0.72
    ibi
    -0.70
    castle
    -0.70
    agra
    -0.69
    POSITIVE LOGITS
     Collider
    0.69
     Aval
    0.69
     largeDownload
    0.68
     Dism
    0.66
     Phant
    0.66
     Reincarn
    0.65
     Cosmos
    0.65
     Pupp
    0.64
    Xi
    0.64
     confuse
    0.63
    Act Density 5.496%

    No Known Activations