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

    Neuronpedia

    INDEX
    Explanations

    various language endings, suffixes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    matic
    0.53
    ，
    0.48
    ↵↵
    0.46
    lıkla
    0.46
    $,
    0.45
     voila
    0.44
    V
    0.44
    eller
    0.44
    OK
    0.43
    igate
    0.43
    POSITIVE LOGITS
    т
    0.84
    ي
    0.82
    ન
    0.75
    us
    0.73
    ق
    0.72
    ные
    0.70
     are
    0.69
    та
    0.68
    ת
    0.68
    é
    0.66
    Act Density 0.399%

    No Known Activations