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

    Neuronpedia

    INDEX
    Explanations

    sh starts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Needless
    0.42
    {}",
    0.39
    磋
    0.39
     infrequently
    0.37
    होस्
    0.36
     côn
    0.35
    ❑
    0.35
    }$}
    0.35
    }*/
    0.35
    Needless
    0.35
    POSITIVE LOGITS
     sh
    0.48
    SH
    0.46
    SCH
    0.46
     SH
    0.46
    SHE
    0.45
    Щ
    0.43
     шей
    0.42
     シェ
    0.41
     ഷ
    0.41
    шем
    0.41
    Act Density 0.033%

    No Known Activations