INDEX
Explanations
already happened or established
New Auto-Interp
Negative Logits
finally
-1.60
still
-1.53
eventually
-1.48
suddenly
-1.48
uniquement
-1.39
finalmente
-1.38
ancora
-1.35
ultimately
-1.34
recently
-1.31
seulement
-1.30
POSITIVE LOGITS
plenty
1.46
enough
1.33
Already
1.31
already
1.25
Already
1.21
some
1.21
several
1.20
before
1.16
sẵn
1.11
established
1.10
Activations Density 0.039%