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

    Neuronpedia

    INDEX
    Explanations

    phrases like "could beat"

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    H
    0.55
     Energ
    0.50
     Architecture
    0.49
    W
    0.49
    ت
    0.49
    Z
    0.48
    c
    0.48
     I
    0.48
    J
    0.48
    t
    0.47
    POSITIVE LOGITS
    них
    0.52
     প্যা
    0.52
    ﺪ
    0.49
     кою
    0.48
    ੁੱ
    0.48
     gelap
    0.46
     чыныгы
    0.45
     패
    0.45
     berdiri
    0.45
    منی
    0.45
    Act Density 0.000%

    No Known Activations