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

    Neuronpedia

    INDEX
    Explanations

    stars twinkling or igniting

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ({\
    0.45
    =
    0.38
    ̿
    0.38
     fermentum
    0.37
    ίσ
    0.37
    一流
    0.36
    )%
    0.36
    emás
    0.36
    力
    0.36
    Starting
    0.36
    POSITIVE LOGITS
    彡
    0.87
     ⭐
    0.80
     twinkle
    0.70
    🌟
    0.70
    ⭐⭐
    0.66
    ⭐
    0.63
    ☆☆
    0.62
     twinkling
    0.62
    璀
    0.61
     aster
    0.60
    Act Density 0.021%

    No Known Activations