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

    Neuronpedia

    INDEX
    Explanations

    writing introductory phrases

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Bitte
    0.74
    壢
    0.71
    veriş
    0.70
    daar
    0.69
    uild
    0.67
    UNDO
    0.66
    HLER
    0.66
     boasts
    0.64
     xác
    0.64
     vudd
    0.64
    POSITIVE LOGITS
     writing
    2.28
     write
    2.20
    Writing
    2.10
     Writing
    2.08
     escribir
    2.00
     Write
    1.99
    Write
    1.95
     wrote
    1.95
     writes
    1.94
     пишу
    1.93
    Act Density 0.091%

    No Known Activations