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

    Neuronpedia

    INDEX
    Explanations

    key followed by topic noun

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    s
    2.03
    i
    1.95
    ের
    1.72
    al
    1.63
    g
    1.62
    c
    1.60
    le
    1.59
    ات
    1.46
    ae
    1.45
    ch
    1.45
    POSITIVE LOGITS
     dotyczą
    1.23
     dừng
    1.22
     profondément
    1.21
    습니다
    1.20
    ﻪ
    1.19
    шая
    1.13
     CHANGES
    1.10
    ći
    1.09
    ницу
    1.08
    迫
    1.07
    Act Density 0.412%

    No Known Activations