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

    Neuronpedia

    INDEX
    Explanations

    introducing an alternative

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    有無
    0.73
     /,
    0.70
    មិន
    0.69
    不易
    0.68
    徨
    0.68
     Jangan
    0.67
    没有什么
    0.66
     không
    0.66
    মান
    0.66
    唔
    0.66
    POSITIVE LOGITS
     instead
    3.08
     Instead
    2.74
    Instead
    2.69
    instead
    2.40
     বরং
    2.07
     invece
    1.93
    而是
    1.88
     вместо
    1.82
    代わりに
    1.77
     rather
    1.72
    Act Density 0.756%

    No Known Activations