INDEX
Explanations
words related to choice or alternatives
New Auto-Interp
Negative Logits
icot
-0.17
elan
-0.16
aji
-0.16
ijd
-0.15
erner
-0.15
anye
-0.15
riangle
-0.15
rup
-0.14
ahi
-0.14
plits
-0.14
POSITIVE LOGITS
close
0.19
closely
0.18
soon
0.18
proxy
0.16
close
0.16
partial
0.16
soon
0.16
à¹ĥà¸ģล
0.15
fake
0.15
near
0.15
Activations Density 0.146%