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

    Neuronpedia

    INDEX
    Explanations

    horizontal and vertical positions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     of
    0.54
     in
    0.48
    ك
    0.45
     قالب
    0.43
    克兰
    0.42
    kost
    0.42
    នៃ
    0.41
    =\$
    0.41
     ofx
    0.39
     Με
    0.39
    POSITIVE LOGITS
     horizontal
    1.17
     Horizontal
    1.07
    Horizontal
    1.02
     vertical
    1.01
    horizontal
    0.95
     Vertical
    0.94
    vertical
    0.90
     verticales
    0.90
     horizontally
    0.89
     horizontale
    0.86
    Act Density 0.024%

    No Known Activations