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

    Neuronpedia

    INDEX
    Explanations

    words related to programming concepts and code structures

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    alties
    -0.71
    uez
    -0.68
    son
    -0.66
    annel
    -0.66
    erred
    -0.66
    lower
    -0.64
    reth
    -0.63
    akh
    -0.63
    anguages
    -0.62
    brow
    -0.62
    POSITIVE LOGITS
     trooper
    0.70
     Yamato
    0.69
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    0.66
     clone
    0.65
    xon
    0.63
     Trooper
    0.60
     Doodle
    0.59
     Skywalker
    0.59
     troopers
    0.57
     clones
    0.57
    Act Density 5.907%

    No Known Activations