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

    Neuronpedia

    INDEX
    Explanations

    section header labelsauthor address TITLE

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     from
    -1.31
    WindowManager
    -0.99
    QMetaType
    -0.98
    ikut
    -0.96
     kø
    -0.95
    १
    -0.94
    resave
    -0.94
    ген
    -0.93
     yang
    -0.92
     maart
    -0.92
    POSITIVE LOGITS
     alcune
    1.18
     accogli
    1.16
     chiunque
    1.08
     piú
    1.07
     possiamo
    1.04
     alcuni
    1.04
     cálido
    1.03
    욤
    1.02
     giusta
    1.02
     eccellente
    1.00
    Act Density 0.007%

    No Known Activations