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

    Neuronpedia

    INDEX
    Explanations

    corpses or wreckage

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    魔法
    0.52
     magic
    0.51
     Magical
    0.51
    神奇
    0.51
     invis
    0.49
     magical
    0.48
     tricky
    0.47
     invisible
    0.47
     mystical
    0.47
     velvety
    0.47
    POSITIVE LOGITS
     corpses
    1.70
     wreckage
    1.69
     cadáver
    1.57
     corpse
    1.49
     carcasses
    1.45
     debris
    1.45
     remnants
    1.42
     wrecked
    1.38
     ruins
    1.30
     carcass
    1.30
    Act Density 0.274%

    No Known Activations