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

    Neuronpedia

    INDEX
    Explanations

    comparing different styles

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    有效
    0.62
    有效的
    0.57
    可靠
    0.50
     useful
    0.48
    Useful
    0.48
     berguna
    0.48
    美味
    0.47
     harmless
    0.47
     baddies
    0.47
     Valuable
    0.47
    POSITIVE LOGITS
     cautious
    0.77
     exploratory
    0.75
     analytical
    0.72
     pragmatic
    0.69
    analytical
    0.66
     emphatic
    0.66
     prescriptive
    0.66
     intros
    0.65
     idealistic
    0.64
     contemplative
    0.64
    Act Density 0.018%

    No Known Activations