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

    Neuronpedia

    INDEX
    Explanations

    variational inference, autoencoders, adversarial

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    一
    0.86
     in
    0.79
    有
    0.79
    с
    0.76
    大
    0.73
     that
    0.71
    在
    0.70
    ”
    0.70
    能
    0.70
    U
    0.69
    POSITIVE LOGITS
     nouvel
    0.86
    owanie
    0.74
     Espanha
    0.69
     roues
    0.68
     Putri
    0.66
    owaniu
    0.66
    )$.
    0.66
     obten
    0.65
    Stencil
    0.65
    (
    0.65
    Act Density 0.003%

    No Known Activations