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

    Neuronpedia

    INDEX
    Explanations

    why things serve a purpose

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Requires
    0.45
    事业
    0.43
    带来
    0.43
     стреми
    0.41
     গিয়েছে
    0.40
     Require
    0.40
    带来了
    0.40
    requires
    0.40
     가져
    0.39
     굉장
    0.39
    POSITIVE LOGITS
     serve
    1.61
     serves
    1.55
     Serve
    1.37
    serve
    1.28
     served
    1.27
    Serve
    1.25
     служит
    1.22
     dient
    1.17
     dienen
    1.16
     sirve
    1.13
    Act Density 0.027%

    No Known Activations