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

    Neuronpedia

    INDEX
    Explanations

    skills and attributes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     redacted
    0.41
     trivially
    0.40
     YAML
    0.40
     trivial
    0.40
     nonempty
    0.39
     unenforceable
    0.39
     lieux
    0.38
     subparagraph
    0.38
    Emoji
    0.38
     explicitly
    0.37
    POSITIVE LOGITS
     defensive
    0.94
     defensively
    0.86
     versatility
    0.84
     athleticism
    0.78
     Defensive
    0.77
     versatile
    0.70
     finesse
    0.67
     physicality
    0.64
     umiejęt
    0.62
     skilled
    0.62
    Act Density 0.035%

    No Known Activations