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

    Neuronpedia

    INDEX
    Explanations

    irregular and improper classifications

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ہ
    0.57
    .
    0.55
     schön
    0.54
     acara
    0.53
    ä
    0.52
    ைகள்
    0.49
     grise
    0.48
    คุณ
    0.48
     mellan
    0.47
     uten
    0.47
    POSITIVE LOGITS
    r
    0.64
    jq
    0.58
     ຈ
    0.58
    irregular
    0.57
    speeds
    0.56
    huang
    0.55
    ।'
    0.54
    pictureBox
    0.53
     Вул
    0.53
    sentences
    0.52
    Act Density 0.003%

    No Known Activations