INDEX
    Explanations
    New Auto-Interp
    Negative Logits
     ej
    -0.08
    -0.08
     нас
    -0.08
     Romantic
    -0.07
    luv
    -0.07
     broaden
    -0.07
     pek
    -0.07
     plateau
    -0.07
     update
    -0.07
     receptive
    -0.07
    POSITIVE LOGITS
     beads
    0.10
     bead
    0.08
     necklaces
    0.08
     Turk
    0.08
     dye
    0.08
    0.08
     сти
    0.08
    0.08
    designer
    0.08
     solides
    0.08
    Act Density 0.003%

    No Known Activations