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

    Neuronpedia

    INDEX
    Explanations

    Can you/I

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     want
    0.70
     quite
    0.68
     crave
    0.66
     inol
    0.64
     often
    0.64
     grim
    0.63
     disag
    0.62
     ultimately
    0.62
     revel
    0.61
     lease
    0.61
    POSITIVE LOGITS
     不能
    0.92
     нельзя
    0.82
    이라고
    0.79
     وهكذا
    0.75
     auxquelles
    0.75
    你們
    0.74
    ister
    0.74
     כך
    0.74
     связи
    0.74
     그리고
    0.73
    Act Density 0.111%

    No Known Activations