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

    Neuronpedia

    INDEX
    Explanations

    rooted in phrases

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     แรง
    0.51
    ajouter
    0.49
     버전
    0.48
     Много
    0.47
     металли
    0.46
    INGLE
    0.45
     ഉപയോഗ
    0.45
    ENGE
    0.45
     హత్య
    0.45
     اخیر
    0.45
    POSITIVE LOGITS
    ↵↵
    0.50
     id
    0.48
     privately
    0.48
     the
    0.45
    <start_of_image>
    0.45
        
    0.44
     N
    0.44
    0.44
     withdrawn
    0.44
    ></
    0.43
    Act Density 0.002%

    No Known Activations