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

    Neuronpedia

    INDEX
    Explanations

    numbersnumbers following 1numbersnumbers in addresses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    er
    3.35
    ০
    3.18
    es
    3.18
    y
    3.04
    en
    3.02
    oretically
    2.97
    ת
    2.96
    i
    2.91
    o
    2.89
    ي
    2.88
    POSITIVE LOGITS
    Й
    1.38
    Согласно
    1.37
    எம்
    1.36
    了一個
    1.29
    Пере
    1.29
    netHex
    1.28
    Вы
    1.24
    Ч
    1.24
    了一
    1.23
    Мы
    1.22
    Act Density 0.535%

    No Known Activations