INDEX
Explanations
high-frequency function words or pronouns in the text
New Auto-Interp
Negative Logits
terra
-0.18
Cherry
-0.15
='".
-0.15
utsch
-0.15
baik
-0.14
udic
-0.14
utters
-0.14
itted
-0.14
_DECLARE
-0.14
Cout
-0.14
POSITIVE LOGITS
åī²
0.16
dbh
0.15
çī
0.15
antics
0.15
allo
0.15
/weather
0.14
JK
0.14
Bundle
0.14
Kant
0.14
Bought
0.14
Activations Density 0.000%