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

    Neuronpedia

    INDEX
    Explanations

    neural network layers

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     vol
    0.70
     எழுத்த
    0.65
     বস্ত্র
    0.65
    kich
    0.62
     अंग्रे
    0.62
    ጳ
    0.61
    uele
    0.61
     cope
    0.61
    bek
    0.60
     گلوکار
    0.60
    POSITIVE LOGITS
     Dense
    0.78
     पाहि
    0.72
     dense
    0.72
    Dense
    0.71
    Dropped
    0.68
     Nang
    0.66
     recounts
    0.65
    ประ
    0.65
    dropped
    0.64
     dropped
    0.64
    Act Density 0.062%

    No Known Activations