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

    Neuronpedia

    INDEX
    Explanations

    correctness and performance evaluation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     yapacağız
    0.49
     garantiert
    0.41
     emphasized
    0.41
     stricter
    0.41
     convertirse
    0.41
     подчер
    0.40
     bestimm
    0.40
    絬
    0.40
     확률
    0.40
     رسمی
    0.39
    POSITIVE LOGITS
     satisfactory
    0.92
     satisfactorily
    0.82
     satisfactor
    0.80
     completeness
    0.77
     demonstrates
    0.75
     appropriateness
    0.75
     adequacy
    0.75
     unsatisfactory
    0.70
     commendable
    0.70
     exceeds
    0.69
    Act Density 0.074%

    No Known Activations