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

    Neuronpedia

    INDEX
    Explanations

    separate digits

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    磷
    0.38
     автобу
    0.38
     маник
    0.36
    🚭
    0.35
    Religious
    0.35
    Luther
    0.35
    hensive
    0.34
    Hector
    0.34
    പാട
    0.34
     сексуа
    0.34
    POSITIVE LOGITS
     separate
    0.46
     Separate
    0.40
     আলাদা
    0.40
     related
    0.38
     hierarchy
    0.38
     Tk
    0.37
     Tz
    0.37
     relating
    0.36
     explicitly
    0.36
     Do
    0.36
    Act Density 0.235%

    No Known Activations