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

    Neuronpedia

    INDEX
    Explanations

    shakes

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     tiros
    0.48
    婼
    0.48
     json
    0.47
     jaga
    0.46
     aí
    0.46
    ftl
    0.45
    Into
    0.45
     ditth
    0.45
     нужно
    0.44
     Into
    0.44
    POSITIVE LOGITS
     him
    0.71
     her
    0.68
    him
    0.56
     ей
    0.55
    her
    0.50
    Her
    0.50
     them
    0.49
    them
    0.49
     his
    0.49
     henne
    0.48
    Act Density 0.001%

    No Known Activations