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

    Neuronpedia

    INDEX
    Explanations

    SQL and code identifiers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     furl
    2.39
    vamo
    2.15
    त
    2.07
     дому
    2.06
    σ
    2.05
    ب
    2.02
    ্সা
    2.00
    醂
    1.90
    зирован
    1.83
    सभी
    1.83
    POSITIVE LOGITS
    ה
    2.76
    া
    2.41
     overpowering
    2.17
    وں
    2.09
    ли
    2.08
    ти
    2.05
     Increasing
    1.91
    ார்க்க
    1.87
    াঃ
    1.87
    டம்
    1.85
    Act Density 0.111%

    No Known Activations