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

    Neuronpedia

    INDEX
    Explanations

    fun and

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    is
    1.66
    o
    1.53
    t
    1.52
     by
    1.52
    el
    1.51
    a
    1.49
    ing
    1.48
    z
    1.46
    e
    1.44
    was
    1.40
    POSITIVE LOGITS
    เป็น
    1.15
     préférable
    1.14
     υψη
    1.09
    จะ
    1.06
     ότι
    1.06
     οπο
    1.06
     дов
    1.06
     ισ
    1.05
     εγκα
    1.04
     ανα
    1.03
    Act Density 2.227%

    No Known Activations