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

    Neuronpedia

    INDEX
    Explanations

    seeing things emerge

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     allows
    0.47
     allowing
    0.46
     indefinitely
    0.45
     considers
    0.45
     Allows
    0.44
     Allowing
    0.43
     undisclosed
    0.42
    เพื่อ
    0.42
     เพื่อ
    0.42
     biar
    0.40
    POSITIVE LOGITS
     firsthand
    0.71
    熟悉的
    0.51
     glimpses
    0.47
     commotion
    0.46
     unfold
    0.44
     apparaître
    0.44
     झलक
    0.43
     glimpse
    0.42
     👀
    0.42
     emerge
    0.41
    Act Density 0.065%

    No Known Activations