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

    Neuronpedia

    INDEX
    Explanations

    Can you

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     there
    -2.80
     to
    -2.33
     as
    -2.19
     if
    -2.19
     for
    -2.13
     I
    -2.03
    );
    -1.86
    There
    -1.77
    };
    -1.76
    for
    -1.72
    POSITIVE LOGITS
    囖
    1.95
    TryDecode
    1.92
    芣
    1.89
    Stap
    1.88
    caneca
    1.81
     .........
    1.77
    1.77
     .......
    1.75
    够了
    1.73
    苤
    1.73
    Act Density 0.006%

    No Known Activations