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

    Neuronpedia

    INDEX
    Explanations

    torch.device("cuda" or "cpu")

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ्स
    0.76
     срок
    0.71
    μμ
    0.71
     uống
    0.71
    RFC
    0.70
     urge
    0.69
    Dental
    0.68
    码
    0.67
     shelled
    0.67
    Roe
    0.66
    POSITIVE LOGITS
    ़
    0.68
    ت
    0.61
    둬
    0.60
     народов
    0.59
    পূর্ণ
    0.57
    ਡ
    0.56
     galore
    0.56
    ট
    0.55
    б
    0.53
    鮒
    0.52
    Act Density 0.003%

    No Known Activations