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

    Neuronpedia

    INDEX
    Explanations

    actions performed by 'he'

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Him
    1.12
     Their
    1.06
     deras
    0.99
     theyre
    0.97
     Worst
    0.94
     mereka
    0.94
     Mr
    0.93
     Laser
    0.92
     Worth
    0.92
     Husband
    0.92
    POSITIVE LOGITS
     himself
    1.55
    his
    1.39
     his
    1.24
     seiner
    1.07
     seine
    1.05
    但他
    1.03
    了他的
    1.02
    他的
    0.95
     мог
    0.91
     kanyang
    0.88
    Act Density 0.016%

    No Known Activations