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

    Neuronpedia

    INDEX
    Explanations

    declining harmful instructions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ดี
    0.44
    好
    0.44
    เมื่อ
    0.42
     dù
    0.42
    저
    0.40
     meski
    0.40
     ดี
    0.38
    മി
    0.37
    winged
    0.37
    ក្រោម
    0.37
    POSITIVE LOGITS
     analyzed
    0.44
    ตำบล
    0.44
    看出
    0.42
     qualitatively
    0.41
    urnal
    0.41
    ponden
    0.40
    dasarkan
    0.40
    0.39
     viewed
    0.39
     rendement
    0.39
    Act Density 0.002%

    No Known Activations