INDEX
Explanations
emotional expressions related to frustration and longing
New Auto-Interp
Negative Logits
"..\..\..\
-0.74
kaynağından
-0.74
UserScript
-0.68
"),
-0.66
تقاوى
-0.61
-0.60
"):
-0.59
"..\..\
-0.59
subcategory
-0.57
}}$}
-0.56
POSITIVE LOGITS
sorry
0.65
please
0.62
Sorry
0.58
̍t
0.58
<eos>
0.57
↵↵
0.55
sorry
0.54
isolato
0.53
Come
0.53
OK
0.53
Activations Density 0.161%