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

    Neuronpedia

    INDEX
    Explanations

    bracketed numbers/placeholders

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    btn
    -0.08
    Bytes
    -0.07
    Spectrum
    -0.07
    Parking
    -0.07
    /t
    -0.07
     promptly
    -0.07
    ARM
    -0.07
    bytes
    -0.07
    Cher
    -0.07
     ਮ
    -0.07
    POSITIVE LOGITS
    راتيجية
    0.09
     дир
    0.08
     toda
    0.08
    ifiques
    0.08
     plantilla
    0.08
     predefined
    0.08
     disciplina
    0.08
    аются
    0.08
    anchors
    0.08
     포함
    0.08
    Act Density 0.014%

    No Known Activations