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

    Neuronpedia

    INDEX
    Explanations

    code function and variable names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    丙基
    0.68
    :>
    0.64
     betrachtet
    0.64
     teilweise
    0.63
    credible
    0.60
    νης
    0.59
    である
    0.59
     proposito
    0.59
     yaitu
    0.59
    ={"
    0.59
    POSITIVE LOGITS
     affectionate
    0.60
     adopting
    0.60
     Би
    0.60
    贊
    0.59
    FormBorderStyle
    0.59
    ಸ್ಕ
    0.58
     treaties
    0.58
     bioge
    0.58
     Non
    0.58
    ٹو
    0.58
    Act Density 0.265%

    No Known Activations