INDEX
Explanations
text with strong emotional or thematic contrasts
New Auto-Interp
Negative Logits
ubbo
-0.16
unkt
-0.15
647
-0.15
asper
-0.15
ihan
-0.15
-validator
-0.14
uffman
-0.14
STRU
-0.14
AppModule
-0.14
Strat
-0.14
POSITIVE LOGITS
Clayton
0.16
пов
0.16
sugar
0.16
Sugar
0.16
Zoom
0.16
Sugar
0.15
ara
0.15
ateurs
0.15
Sug
0.15
Strict
0.14
Activations Density 0.026%