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

    Neuronpedia

    INDEX
    Explanations

    class inheritance (`extends` or `class(...)`)

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ，
    0.78
    бна
    0.66
     tới
    0.65
     (
    0.64
    нения
    0.61
    чному
    0.61
     når
    0.60
    ния
    0.60
    ünden
    0.59
    ،
    0.59
    POSITIVE LOGITS
    in
    1.26
    a
    0.85
    w
    0.84
     in
    0.82
    f
    0.71
     had
    0.70
     has
    0.68
    d
    0.65
    um
    0.65
    el
    0.64
    Act Density 0.397%

    No Known Activations