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

    Neuronpedia

    INDEX
    Explanations

    occlusion

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    闐
    -0.98
    のではない
    -0.98
    zsche
    -0.96
     lørdag
    -0.94
     siglas
    -0.94
    ccedil
    -0.93
    蕤
    -0.92
    Ostat
    -0.91
     ajoute
    -0.90
    ","\
    -0.90
    POSITIVE LOGITS
    初心
    1.02
     &
    0.95
     \&
    0.93
     this
    0.92
     |
    0.87
     locks
    0.85
     automatically
    0.84
     -
    0.84
     dr
    0.84
     взял
    0.84
    Act Density 0.002%

    No Known Activations