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

    Neuronpedia

    INDEX
    Explanations

    shifting

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     accents
    0.63
     accompanying
    0.63
     competing
    0.61
     gravid
    0.60
     tiger
    0.60
     laid
    0.59
     corresponding
    0.59
     favored
    0.58
     car
    0.58
     last
    0.57
    POSITIVE LOGITS
    GULD
    0.82
    േഖ
    0.80
    akespeare
    0.76
    ahr
    0.74
     там
    0.74
    riek
    0.74
    ⤥
    0.73
    okedex
    0.73
     dív
    0.72
     నిర్
    0.72
    Act Density 0.053%

    No Known Activations