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

    Neuronpedia

    INDEX
    Explanations

    surprisingly complex

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     اكيد
    0.49
    แน่นอน
    0.48
     tentunya
    0.46
     évidemment
    0.46
     sicherlich
    0.43
     obviously
    0.43
    겠지만
    0.42
     sicuramente
    0.42
     নিঃসন্দেহে
    0.42
     oczywiście
    0.41
    POSITIVE LOGITS
     surprisingly
    1.11
    Surprisingly
    0.95
     Surprisingly
    0.91
    竟然
    0.88
    surprisingly
    0.86
     actually
    0.81
    居然
    0.79
    actually
    0.79
     oddly
    0.78
     strangely
    0.76
    Act Density 0.075%

    No Known Activations