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

    Neuronpedia

    INDEX
    Explanations

    matrix decomposition

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    aito
    0.83
     バ
    0.77
     তাত
    0.72
    tc
    0.71
     Schlüssel
    0.71
     キー
    0.70
    atonin
    0.70
     клавиа
    0.70
     außerdem
    0.70
     chữ
    0.69
    POSITIVE LOGITS
     Relating
    0.79
    meyen
    0.79
    きたいと思います
    0.78
     preserving
    0.77
    regarding
    0.75
    面向
    0.74
    Dried
    0.74
    relating
    0.73
     restoring
    0.69
     proposing
    0.68
    Act Density 0.005%

    No Known Activations