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

    Neuronpedia

    INDEX
    Explanations

    I'll do my best

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    具体
    0.67
    bre
    0.62
    br
    0.57
    la
    0.56
    平衡
    0.53
     Specifically
    0.53
    কু
    0.53
    bl
    0.52
    安全性
    0.51
    Specifically
    0.51
    POSITIVE LOGITS
     best
    1.29
     Best
    1.06
    best
    1.00
     BEST
    0.98
    Best
    0.98
    BEST
    0.93
     terbaik
    0.92
     নিঃসৃত
    0.89
     clerical
    0.89
     내용
    0.87
    Act Density 0.082%

    No Known Activations