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

    Neuronpedia

    INDEX
    Explanations

    high value items

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    j
    0.82
    ون
    0.79
     in
    0.78
    מ
    0.74
    )
    0.71
    м
    0.69
    os
    0.68
     be
    0.67
     five
    0.67
    ก
    0.66
    POSITIVE LOGITS
    cyclic
    0.65
    <unused444>
    0.64
    𖥔
    0.60
    I
    0.59
     matem
    0.58
    acetyl
    0.58
    Athlete
    0.58
    пуль
    0.57
    aginaw
    0.56
    beza
    0.56
    Act Density 0.003%

    No Known Activations