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

    Neuronpedia

    INDEX
    Explanations

    good, efficient, easy, well

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     potrebbe
    0.59
    😖
    0.55
     möglicherweise
    0.54
     pourraient
    0.54
     pourrait
    0.52
     potrebbero
    0.52
     อาจ
    0.51
     moeten
    0.49
     могла
    0.48
     podrían
    0.48
    POSITIVE LOGITS
     excellently
    0.88
     beautifully
    0.86
     thankfully
    0.82
     plenty
    0.81
     wonderfully
    0.80
     superbly
    0.80
     fortunately
    0.77
     ample
    0.76
     прекрасно
    0.75
     brilliantly
    0.74
    Act Density 0.186%

    No Known Activations