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

    Neuronpedia

    INDEX
    Explanations

    best or ideal option

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Possible
    0.44
     Possible
    0.43
     можли
    0.41
    可能
    0.40
    有名な
    0.36
    possible
    0.35
     возможных
    0.35
    最後に
    0.35
     מאוד
    0.35
     возмо
    0.34
    POSITIVE LOGITS
     preferred
    0.87
     ideal
    0.77
     way
    0.72
    preferred
    0.69
     preferable
    0.68
     correct
    0.65
     superior
    0.64
     recommended
    0.63
     best
    0.61
     идеа
    0.60
    Act Density 0.076%

    No Known Activations