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

    Neuronpedia

    INDEX
    Explanations

    ending punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     azonban
    0.90
     כאשר
    0.90
     सर्वप्रथम
    0.88
     однако
    0.84
     jedoch
    0.83
    具有
    0.81
     различными
    0.80
    较为
    0.80
    由于
    0.79
    注意
    0.79
    POSITIVE LOGITS
     Seriously
    1.55
    Seriously
    1.33
     Gotta
    1.20
     Forget
    1.16
     Literally
    1.12
     Yeah
    1.10
     Get
    1.09
     Think
    1.07
     Really
    1.07
     Yep
    1.07
    Act Density 1.186%

    No Known Activations