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

    Neuronpedia

    INDEX
    Explanations

    non-English characters and symbols, particularly those related to Latin and Cyrillic scripts

    oai_token-act-pair · gpt-3.5-turbo

    characters or symbols used in encoding or formatting, possibly in a specific language or script

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    olphin
    -0.73
    swick
    -0.72
    nesday
    -0.72
    ifference
    -0.70
    aday
    -0.67
    WARD
    -0.67
    essa
    -0.67
     externalToEVAOnly
    -0.65
    eryl
    -0.63
     Cycling
    -0.62
    POSITIVE LOGITS
    ¾
    1.56
    Į
    1.54
    ©
    1.49
    Ĵ
    1.49
    ¼
    1.46
    Ķ
    1.46
    ¸
    1.42
    ĥ
    1.41
    ħ
    1.41
    ĩ
    1.36
    Act Density 0.011%

    No Known Activations