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

    Neuronpedia

    INDEX
    Explanations

    It appears that Neuron 4 does not show any activation for the given inputs, suggesting that it is not finding the specific feature or pattern it is designed to detect within this text. Therefore, there is no activation or feature identified by Neuron 4 in the provided samples

    oai_token-act-pair · gpt-4-turbo
    New Auto-Interp
    Negative Logits
     sqor
    -0.80
     seams
    -0.71
     eyed
    -0.68
    oshenko
    -0.65
     glim
    -0.64
     untrue
    -0.63
     demos
    -0.63
     shader
    -0.63
     Witches
    -0.62
     pse
    -0.62
    POSITIVE LOGITS
    ¥µ
    0.66
    rive
    0.66
    utm
    0.64
    iday
    0.63
    ér
    0.62
    rug
    0.61
    spection
    0.61
     CrossRef
    0.60
    ij士
    0.60
    athing
    0.60
    Act Density 0.000%

    No Known Activations

    This feature has no known activations.