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

    Neuronpedia

    INDEX
    Explanations

    fixed-width monospace font

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Dynamic
    0.58
     dynamic
    0.56
     динами
    0.55
    Dynamic
    0.54
     code
    0.52
     Code
    0.52
    动态
    0.50
    dynamic
    0.49
    code
    0.48
    的代码
    0.46
    POSITIVE LOGITS
     Fixed
    0.79
     fixed
    0.77
    Fixed
    0.77
    固定
    0.71
     FIXED
    0.68
    fixed
    0.65
     固定
    0.63
     fixe
    0.55
     fixa
    0.55
     фикси
    0.55
    Act Density 0.005%

    No Known Activations