INDEX
Explanations
phrases where the phrase "In fact," is present
phrases that emphasize certainty or affirmation
New Auto-Interp
Negative Logits
battle
-0.65
thal
-0.60
owed
-0.59
atlantic
-0.59
que
-0.59
crunch
-0.58
gment
-0.58
afi
-0.58
alysed
-0.57
airs
-0.57
POSITIVE LOGITS
according
0.85
although
0.81
though
0.77
however
0.76
contrary
0.76
despite
0.73
unlike
0.72
interestingly
0.67
Restaur
0.67
notwithstanding
0.67
Activations Density 0.092%