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

    Neuronpedia

    INDEX
    Explanations

    object attributes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ह
    1.98
    ন
    1.96
    ما
    1.83
    િંગ
    1.82
    ر
    1.76
    ός
    1.73
     ifs
    1.70
     тех
    1.66
     lua
    1.66
    بر
    1.62
    POSITIVE LOGITS
    dan
    2.23
    it
    2.06
     лишь
    2.02
    tar
    2.00
    FAs
    1.99
     рассказал
    1.95
    vitro
    1.95
    dal
    1.94
    ம்
    1.93
     meteen
    1.93
    Act Density 0.069%

    No Known Activations