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

    Neuronpedia

    INDEX
    Explanations

    early detection of problems

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     $_{
    0.44
     $_{\
    0.44
     criterio
    0.43
    。
    0.40
    zeichnis
    0.40
     丶
    0.39
     然而
    0.38
    ின்றன
    0.38
     ഇത്
    0.38
    می
    0.38
    POSITIVE LOGITS
    4
    0.49
    ιος
    0.46
     wsi
    0.42
    pdfs
    0.42
    pann
    0.41
    ٽ
    0.41
     offsetting
    0.41
    6
    0.40
     restructured
    0.40
    <unused62>
    0.39
    Act Density 0.003%

    No Known Activations