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

    Neuronpedia

    INDEX
    Explanations

    shade of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ى
    1.92
    м
    1.91
    р
    1.88
    ี
    1.78
    не
    1.69
    нда
    1.66
    ु
    1.66
    ра
    1.63
    ро
    1.60
    ्
    1.60
    POSITIVE LOGITS
     phố
    2.17
     особое
    2.02
    ის
    2.00
    Жи
    2.00
     bruto
    1.95
     intim
    1.94
    ılarak
    1.92
    માં
    1.91
     reduz
    1.90
    }$,
    1.89
    Act Density 0.033%

    No Known Activations