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

    Neuronpedia

    INDEX
    Explanations

    year, number, page, edition

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    l
    0.70
     ><
    0.70
     returning
    0.69
    ac
    0.67
    m
    0.67
    return
    0.67
     dan
    0.64
     using
    0.64
    is
    0.63
    are
    0.63
    POSITIVE LOGITS
    <unused394>
    1.01
     Reprint
    0.93
    <unused412>
    0.92
    <unused480>
    0.92
    <unused1930>
    0.89
    <unused410>
    0.87
    <unused2094>
    0.86
    <unused938>
    0.86
    <unused2078>
    0.85
    <unused1679>
    0.84
    Act Density 0.001%

    No Known Activations