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

    Neuronpedia

    INDEX
    Explanations

    transformer features

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Whether
    0.76
    If
    0.75
    τσ
    0.75
    Usually
    0.75
    Using
    0.74
    Which
    0.74
    This
    0.73
    Your
    0.72
    Does
    0.71
    MSE
    0.70
    POSITIVE LOGITS
     transforma
    0.81
     emergent
    0.80
     combinatorial
    0.79
     paradig
    0.77
     multidiscipl
    0.77
     ಸಾಮ
    0.75
     variegated
    0.74
    cedented
    0.74
    segaretro
    0.74
    êmes
    0.74
    Act Density 0.268%

    No Known Activations