INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    	project
    -0.07
    Acc
    -0.07
    .dir
    -0.07
    UE
    -0.07
    Song
    -0.07
    UR
    -0.06
    _female
    -0.06
    QUAL
    -0.06
     rum
    -0.06
    文化
    -0.06
    POSITIVE LOGITS
    -Saharan
    0.07
     MCP
    0.07
     İzmir
    0.06
     соот
    0.06
     Shut
    0.06
    0.06
     dayan
    0.06
    yling
    0.06
    TransparentColor
    0.06
     bois
    0.06
    Act Density 0.152%

    No Known Activations