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

    Neuronpedia

    INDEX
    Explanations

    matrix compatibility and multiplication

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     sentimento
    -0.09
     constructive
    -0.08
     thief
    -0.08
    goto
    -0.08
    öse
    -0.08
     literal
    -0.08
     Ра
    -0.07
    onav
    -0.07
     goto
    -0.07
     генерал
    -0.07
    POSITIVE LOGITS
     mismatch
    0.15
    尺寸
    0.15
    Mismatch
    0.14
     mism
    0.14
     compatibility
    0.13
     taille
    0.13
     incompat
    0.13
     tamaño
    0.13
    サイズ
    0.13
     incompatible
    0.13
    Act Density 0.008%

    No Known Activations