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

    Neuronpedia

    INDEX
    Explanations

    _line

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    (static
    -0.07
     aloud
    -0.07
    "He
    -0.07
    啕
    -0.07
     mét
    -0.07
    Adult
    -0.07
     advertised
    -0.07
    -has
    -0.06
    uddenly
    -0.06
    (Art
    -0.06
    POSITIVE LOGITS
    ем
    0.07
    ש
    0.07
     Insights
    0.07
     checkpoint
    0.07
     efficient
    0.07
     bearings
    0.07
    ndon
    0.07
    בות
    0.07
    press
    0.07
     Büyük
    0.06
    Act Density 0.002%

    No Known Activations