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

    Neuronpedia

    INDEX
    Explanations

    Russian adjective endings

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     odgovor
    -0.09
     Kass
    -0.08
     american
    -0.08
    牡
    -0.08
     Su
    -0.08
    �
    -0.08
    SG
    -0.07
     cáo
    -0.07
     Ful
    -0.07
    按
    -0.07
    POSITIVE LOGITS
     xt
    0.08
     quasi
    0.07
    rema
    0.07
     charter
    0.07
     зат
    0.07
     midfield
    0.07
     pinned
    0.07
    726
    0.07
     gcd
    0.07
    182
    0.07
    Act Density 0.036%

    No Known Activations