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

    Neuronpedia

    INDEX
    Explanations

    choosing among options

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    娀
    0.45
    谀
    0.45
    发言
    0.42
     аласыз
    0.42
     improb
    0.41
     женщина
    0.41
    等地
    0.41
    等は
    0.41
     apar
    0.40
     episcopal
    0.40
    POSITIVE LOGITS
     without
    0.61
     namely
    0.61
     via
    0.61
     whose
    0.60
     with
    0.57
    without
    0.57
    using
    0.56
    with
    0.56
    via
    0.55
     comprising
    0.55
    Act Density 0.012%

    No Known Activations