INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     mantle
    -0.08
     fehl
    -0.08
     Joachim
    -0.08
     MOS
    -0.07
    fas
    -0.07
     vigil
    -0.07
    Proposal
    -0.07
    flight
    -0.07
     Fran
    -0.07
     folos
    -0.07
    POSITIVE LOGITS
    0.10
     терап
    0.08
    0.08
     Berry
    0.08
    -speaking
    0.08
    0.08
     hedge
    0.08
    geving
    0.08
    znym
    0.08
     инвест
    0.08
    Act Density 0.004%

    No Known Activations