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

    Neuronpedia

    INDEX
    Explanations

    prioritize safety, society, efficiency

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     closer
    0.38
    andelion
    0.37
    nul
    0.37
     rings
    0.37
     Bast
    0.37
     through
    0.37
     embracing
    0.37
     new
    0.35
    Reception
    0.35
     Reception
    0.35
    POSITIVE LOGITS
     sacrificed
    1.05
     sacrifice
    1.01
     sacrificing
    0.98
     приорите
    0.94
     prioritizing
    0.91
     sacrific
    0.90
     prioritization
    0.90
    牺牲
    0.90
     sacr
    0.88
     prioritized
    0.88
    Act Density 0.465%

    No Known Activations