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

    Neuronpedia

    INDEX
    Explanations

    negative contractions such as "can't" and "don't."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    .Apis
    -0.15
    inke
    -0.15
    bero
    -0.14
     Bags
    -0.14
    uten
    -0.14
    ors
    -0.14
    fac
    -0.13
     ZERO
    -0.13
    ós
    -0.13
    èĩ´
    -0.13
    POSITIVE LOGITS
     Tang
    0.16
    &page
    0.15
    RU
    0.15
    zag
    0.15
    ((&
    0.14
    amentals
    0.14
    egra
    0.14
    ropa
    0.14
    ibold
    0.14
    Advisor
    0.14
    Act Density 0.030%

    No Known Activations