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

    Neuronpedia

    INDEX
    Explanations

    output in JSON format

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ä
    2.95
    ές
    2.80
     segar
    2.78
     цве
    2.71
    ның
    2.65
    сти
    2.60
     Перейти
    2.60
    どうぞ
    2.60
    的网络
    2.59
    ভাবে
    2.55
    POSITIVE LOGITS
    ു
    3.93
    ت
    3.91
    ::::::::::::::::
    3.89
    :::
    3.57
    g
    3.51
    ::::
    3.42
    го
    3.37
    dehyde
    3.36
    ::::::::
    3.25
    ::::::
    3.19
    Act Density 0.112%

    No Known Activations