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

    Neuronpedia

    INDEX
    Explanations

    source code

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     permanently
    -0.07
    [str
    -0.07
    ,u
    -0.07
    anger
    -0.06
     acquiring
    -0.06
     Johnston
    -0.06
    ustomed
    -0.06
    eware
    -0.06
    วาง
    -0.06
    oul
    -0.06
    POSITIVE LOGITS
    μπο
    0.07
    (mat
    0.06
     verileri
    0.06
    _KeyPress
    0.06
     condol
    0.06
     Deniz
    0.06
    _venta
    0.06
     Savaş
    0.06
     //$
    0.06
    characters
    0.06
    Act Density 0.000%

    No Known Activations