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

    Neuronpedia

    INDEX
    Explanations

    act as a role

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    -
    1.82
    ä
    1.63
     to
    1.49
    ait
    1.37
    og
    1.36
    ل
    1.33
    at
    1.30
    ästä
    1.24
    ow
    1.14
    ва
    1.13
    POSITIVE LOGITS
    t
    1.30
    ின
    1.01
    p
    0.99
     преми
    0.98
    sand
    0.95
    。",
    0.94
    沒
    0.92
    ak
    0.91
    the
    0.89
    shells
    0.88
    Act Density 0.165%

    No Known Activations