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

    Neuronpedia

    INDEX
    Explanations

    preference/preferences

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    와의
    2.27
    𝑝
    1.95
    <0xB2>
    1.89
    没有
    1.85
    一個
    1.84
    fh
    1.84
    o
    1.81
    hearted
    1.80
    로
    1.80
    f
    1.77
    POSITIVE LOGITS
     rutas
    2.14
     sexuales
    2.13
     atentos
    2.08
     útiles
    1.99
    你了
    1.98
    Теперь
    1.95
    Доброго
    1.93
     compuestos
    1.91
    ჲ
    1.90
     loudness
    1.86
    Act Density 0.014%

    No Known Activations