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

    Neuronpedia

    INDEX
    Explanations

    express affection

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     nice
    0.94
    Nice
    0.91
     interesting
    0.91
    Interesting
    0.91
    興味
    0.90
     disagreeable
    0.89
     enjoyable
    0.89
    不错的
    0.87
     Interesting
    0.87
    interesting
    0.87
    POSITIVE LOGITS
     adore
    1.34
     LOVE
    1.31
     love
    1.28
     absolutely
    1.23
     instantly
    1.20
     immediately
    1.12
    LOVE
    1.11
    love
    1.10
     lov
    1.09
     adored
    1.08
    Act Density 0.242%

    No Known Activations