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

    Neuronpedia

    INDEX
    Explanations

    reinforcing facts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     nhưng
    0.55
     لیکن
    0.52
     néanmoins
    0.51
     однако
    0.50
     tuttavia
    0.47
     jedoch
    0.47
     Однако
    0.47
     แต่
    0.47
     pero
    0.46
     แน
    0.44
    POSITIVE LOGITS
     actually
    1.52
     sogar
    1.41
    むしろ
    1.27
     addirittura
    1.26
    actually
    1.24
     zelfs
    1.24
     faktisk
    1.18
    Actually
    1.17
    甚至
    1.16
     Actually
    1.16
    Act Density 0.074%

    No Known Activations