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

    Neuronpedia

    INDEX
    Explanations

    raises an eyebrow or awareness

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    好
    1.30
    ко
    1.16
    يل
    1.13
    。
    1.13
    ！
    1.13
    知道
    1.12
    ા
    1.11
    さん
    1.10
    聞
    1.09
    ام
    1.08
    POSITIVE LOGITS
    ra
    1.55
    ’
    1.52
    al
    1.44
    ot
    1.42
    ores
    1.23
    ir
    1.14
    iation
    1.07
    ol
    1.05
    or
    1.03
    oo
    1.02
    Act Density 0.001%

    No Known Activations