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

    Neuronpedia

    INDEX
    Explanations

    resources that help

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    真正的
    0.79
     самые
    0.76
     própri
    0.74
     সবচেয়ে
    0.73
     최고의
    0.72
     শুধু
    0.70
     সবচেয়ে
    0.70
     ovviamente
    0.70
    最高的
    0.70
     epitome
    0.69
    POSITIVE LOGITS
     helpful
    2.50
     Helpful
    2.19
     useful
    2.18
    helpful
    2.14
     Useful
    1.95
    Useful
    1.88
     도움이
    1.86
    useful
    1.82
     might
    1.80
     help
    1.77
    Act Density 0.702%

    No Known Activations