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

    Neuronpedia

    INDEX
    Explanations

    selecting rows

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Edge
    1.29
     edge
    1.28
     Edge
    1.20
    edge
    1.19
    EDGE
    1.14
     EDGE
    1.08
    边缘
    0.99
    Edges
    0.94
     edges
    0.91
    edges
    0.79
    POSITIVE LOGITS
     rows
    2.33
     row
    2.25
    rows
    2.02
    Rows
    1.95
    row
    1.90
     Row
    1.90
    Row
    1.88
     Rows
    1.85
    ROW
    1.56
     ROW
    1.52
    Act Density 0.204%

    No Known Activations