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

    Neuronpedia

    INDEX
    Explanations

    units or quantities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     is
    0.36
     あ
    0.29
     Oath
    0.28
    ಿಗಳು
    0.28
     indah
    0.27
    ヴァ
    0.27
    無い
    0.27
    کم
    0.27
     biti
    0.27
    isBlank
    0.27
    POSITIVE LOGITS
    at
    0.36
    as
    0.36
     consommateurs
    0.33
    et
    0.32
    ента
    0.32
    on
    0.31
     você
    0.31
     bạn
    0.30
     зве
    0.30
    ر
    0.30
    Act Density 0.988%

    No Known Activations