INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    Hey
    -0.08
     cant
    -0.07
     action
    -0.07
    -basic
    -0.07
     hi
    -0.07
    			
    -0.07
    (seed
    -0.07
    725
    -0.07
    -0.07
     hir
    -0.07
    POSITIVE LOGITS
     refiere
    0.12
     означает
    0.11
    意味着
    0.11
     मतलब
    0.11
     referring
    0.10
     nghĩa
    0.10
     refer
    0.10
     bedeutet
    0.10
    意味
    0.10
     refers
    0.09
    Act Density 0.064%

    No Known Activations