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

    Neuronpedia

    INDEX
    Explanations

    data array changes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    te
    0.90
     the
    0.86
    та
    0.86
    ara
    0.86
    in
    0.85
    im
    0.84
    il
    0.82
    os
    0.80
    ä
    0.80
    our
    0.79
    POSITIVE LOGITS
    ﺖ
    1.01
    ﺪ
    0.97
     dowol
    0.85
     použ
    0.84
     utilizz
    0.82
    ﻮ
    0.81
    ﻒ
    0.80
    ﺐ
    0.80
    ﺮ
    0.79
     sencillo
    0.79
    Act Density 0.782%

    No Known Activations