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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "backwards" in a text

    oai_token-act-pair · gpt-3.5-turbo

    the concept of "backwards" or actions performed in reverse

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    raltar
    -0.74
    "},"
    -0.72
    AIN
    -0.70
    illary
    -0.70
    riz
    -0.68
    ULE
    -0.65
    ittee
    -0.64
    ========
    -0.63
    tein
    -0.63
    te
    -0.61
    POSITIVE LOGITS
    stairs
    0.93
     backward
    0.90
    actively
    0.90
    gradation
    0.89
    ward
    0.86
    Ħ¢
    0.84
    wards
    0.84
    dash
    0.83
     compat
    0.79
     backwards
    0.77
    Act Density 0.006%

    No Known Activations