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

    Neuronpedia

    INDEX
    Explanations

    defining inputs, scope, and terms

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     with
    1.36
     +
    1.34
    t
    1.32
     поги
    1.30
     to
    1.24
     ,
    1.23
     <
    1.21
     =
    1.20
     ولكن
    1.19
     >
    1.16
    POSITIVE LOGITS
    is
    1.99
    in
    1.94
    it
    1.80
    an
    1.63
    as
    1.63
    un
    1.51
    ம்
    1.50
    ar
    1.37
    ad
    1.36
    ו
    1.33
    Act Density 0.074%

    No Known Activations