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

    Neuronpedia

    INDEX
    Explanations

    states and actions, like disengaging

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    isers
    0.44
    öy
    0.44
    ције
    0.42
    lings
    0.42
     wineries
    0.41
    γωγ
    0.41
     устройства
    0.41
     gamle
    0.41
    лі
    0.40
    ért
    0.40
    POSITIVE LOGITS
     효율
    0.45
    3
    0.42
    它可以
    0.39
     violence
    0.39
    ጤ
    0.39
     ৩৪
    0.38
     नेत्र
    0.37
    ほか
    0.37
     जिसकी
    0.37
    yeah
    0.37
    Act Density 0.006%

    No Known Activations