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

    Neuronpedia

    INDEX
    Explanations

    though, or, and

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     want
    0.43
    </h4>
    0.41
     quero
    0.40
     questi
    0.39
     lote
    0.39
     slightest
    0.38
     winkel
    0.38
     produire
    0.37
     nincs
    0.37
     simplest
    0.37
    POSITIVE LOGITS
    ซึ่ง
    0.60
    which
    0.53
    !),
    0.52
    including
    0.49
     désormais
    0.46
     এবং
    0.45
    했고
    0.45
     ซึ่ง
    0.45
     übrigens
    0.45
     whose
    0.44
    Act Density 0.170%

    No Known Activations