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

    Neuronpedia

    INDEX
    Explanations

    colors

    oai_token-act-pair · gpt-3.5-turbo

    colors and their attributes in descriptions of creatures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ========
    -0.84
    ablishment
    -0.83
    =-=-=-=-
    -0.82
    ovie
    -0.75
    ISTORY
    -0.74
    ernel
    -0.73
    ĥ
    -0.72
    Ĩ
    -0.72
    zsche
    -0.72
    rent
    -0.71
    POSITIVE LOGITS
     dye
    1.21
     hue
    1.18
     stripe
    1.09
     colored
    1.09
    prints
    1.08
     stripes
    1.08
     coloured
    1.07
     tint
    1.07
     color
    1.07
     striped
    1.07
    Act Density 0.098%

    No Known Activations