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

    Neuronpedia

    INDEX
    Explanations

    explanations and improvements

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     manchmal
    0.75
     qualche
    0.74
     примерно
    0.66
     alguno
    0.65
     sometimes
    0.64
     atau
    0.64
     ungef
    0.62
    いずれ
    0.61
     substituir
    0.61
     soms
    0.61
    POSITIVE LOGITS
    allows
    1.08
     allows
    1.07
    Allows
    1.06
    allowing
    1.01
     Allows
    0.96
     позволяет
    0.95
     позволяют
    0.95
     permite
    0.91
     allowing
    0.91
     позволя
    0.88
    Act Density 0.539%

    No Known Activations