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

    Neuronpedia

    INDEX
    Explanations

    discussing ethical relationships and transformation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Romantic
    0.48
    romantic
    0.41
    💏
    0.40
     Romantic
    0.39
     legality
    0.39
     romantic
    0.39
     porn
    0.39
    合法
    0.39
    Paths
    0.38
     bloodshed
    0.38
    POSITIVE LOGITS
     overcoming
    0.64
     overcame
    0.63
     overcome
    0.60
     overcomes
    0.56
    克服
    0.55
     positive
    0.55
     interesting
    0.53
     преодо
    0.51
     friendly
    0.50
     solving
    0.50
    Act Density 0.122%

    No Known Activations