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

    Neuronpedia

    INDEX
    Explanations

    blanket &

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ak
    1.80
    nın
    1.76
    n
    1.75
    }$
    1.63
    ia
    1.56
    та
    1.56
    }=\
    1.55
    na
    1.49
    s
    1.48
    op
    1.47
    POSITIVE LOGITS
     takers
    1.45
    선
    1.45
    ized
    1.41
     CZ
    1.41
    ties
    1.38
    هد
    1.38
    ној
    1.36
    ア
    1.36
    いう
    1.34
    就知道
    1.33
    Act Density 0.738%

    No Known Activations