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

    Neuronpedia

    INDEX
    Explanations

    do not, don't, never

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     lacking
    0.37
    धाना
    0.37
    缺乏
    0.37
     Willing
    0.37
     manca
    0.37
     faltando
    0.35
     உண்மைய
    0.35
     부족
    0.35
     throughput
    0.35
     трудно
    0.35
    POSITIVE LOGITS
     jangan
    2.08
    不要
    1.91
     đừng
    1.78
     Jangan
    1.75
    Jangan
    1.73
     不要
    1.69
    jangan
    1.61
    อย่า
    1.56
     อย่า
    1.54
     never
    1.47
    Act Density 0.045%

    No Known Activations