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

    Neuronpedia

    INDEX
    Explanations

    start of code blocks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    mada
    0.40
    瓴
    0.37
    പാട
    0.36
     адам
    0.36
    leye
    0.35
    Allister
    0.35
     Apar
    0.34
    Sullivan
    0.34
    мили
    0.34
    ।]
    0.33
    POSITIVE LOGITS
     temporary
    0.59
     Begin
    0.56
     begin
    0.55
     开始
    0.53
    开始
    0.52
     temporarily
    0.50
     شروع
    0.50
     began
    0.49
     begins
    0.49
     beginnen
    0.49
    Act Density 0.187%

    No Known Activations