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

    Neuronpedia

    INDEX
    Explanations

    unsigned short

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     are
    -2.81
    曁
    -2.00
     после
    -1.93
     )
    -1.91
     neuen
    -1.88
     is
    -1.80
    ↵
    -1.80
    arthur
    -1.76
     larges
    -1.73
     }
    -1.73
    POSITIVE LOGITS
    '
    2.31
    \
    2.09
     ſta
    2.06
     ſy
    2.03
    厹
    2.02
     climático
    1.98
    汌
    1.94
     缝
    1.87
    てて
    1.87
    퐿
    1.86
    Act Density 0.003%

    No Known Activations