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

    Neuronpedia

    INDEX
    Explanations

    titles and abstracts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     oku
    0.47
     malzem
    0.47
     kuts
    0.45
     ~(~)
    0.43
     lounges
    0.42
     haters
    0.42
     échantillons
    0.42
     生地
    0.41
     当然
    0.41
     sanciones
    0.41
    POSITIVE LOGITS
    Evaluating
    0.70
    Using
    0.69
    Improving
    0.68
     Improving
    0.67
     Evaluating
    0.67
     Efficient
    0.67
    Efficient
    0.66
     Comparative
    0.65
    Exploring
    0.64
    Improved
    0.63
    Act Density 0.008%

    No Known Activations