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

    Neuronpedia

    INDEX
    Explanations

    counts and repetitions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    تمد
    0.62
     приро
    0.62
    ựa
    0.61
     blending
    0.60
    当然
    0.59
    பெ
    0.58
     研究
    0.58
    ához
    0.57
     yǔ
    0.57
    hiana
    0.57
    POSITIVE LOGITS
     twice
    3.86
    twice
    3.40
     Twice
    3.33
     repeatedly
    3.28
    多次
    3.17
     thrice
    2.98
    两次
    2.94
     repeated
    2.94
    几次
    2.90
    repeated
    2.88
    Act Density 0.334%

    No Known Activations