INDEX
Explanations
phrases related to value judgments, where things are deemed worth it, genuine, or significant
expressions related to value assessment and significance
New Auto-Interp
Negative Logits
but
-1.04
But
-0.82
But
-0.77
but
-0.76
BUT
-0.71
However
-0.71
However
-0.68
eatured
-0.68
}}
-0.66
BUT
-0.65
POSITIVE LOGITS
nonetheless
2.11
anyway
1.43
nevertheless
1.32
anyways
1.27
etheless
0.97
insofar
0.80
thanks
0.76
owing
0.75
because
0.73
awfully
0.72
Activations Density 0.995%