INDEX
Explanations
verbs and their forms that indicate ongoing actions or states
New Auto-Interp
Negative Logits
strup
-0.22
cheid
-0.18
ilim
-0.17
#
-0.16
ampo
-0.15
717
-0.15
×ķ
-0.15
174
-0.14
leurs
-0.14
afil
-0.14
POSITIVE LOGITS
inals
0.16
Oriental
0.15
rag
0.15
shed
0.15
"),"
0.14
ary
0.14
lah
0.14
id
0.13
çļĦæĺ¯
0.13
consts
0.13
Activations Density 0.287%