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

    Neuronpedia

    INDEX
    Explanations

    evaluating honesty and quantity

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     (€
    0.42
     Elys
    0.39
     ['$
    0.38
     ($
    0.38
     occ
    0.38
     ÷
    0.38
     దగ్
    0.36
     طب
    0.35
     immediately
    0.34
     forfeit
    0.34
    POSITIVE LOGITS
    quinazoline
    0.41
     কিন
    0.39
    绺
    0.38
     পিঁপড়ার
    0.38
    ।--
    0.38
     egg
    0.38
    蛋
    0.37
    луч
    0.36
    鑣
    0.36
    パクト
    0.36
    Act Density 0.004%

    No Known Activations