INDEX
Explanations
emotional connections and personal significance related to experiences and places
New Auto-Interp
Negative Logits
onis
-0.16
omo
-0.16
iku
-0.15
ika
-0.14
adaki
-0.13
idal
-0.13
iek
-0.13
anz
-0.13
isman
-0.13
itty
-0.13
POSITIVE LOGITS
rob
0.14
.asp
0.14
_endian
0.14
imd
0.13
669
0.13
.blog
0.13
prime
0.13
ibraltar
0.13
334
0.13
ple
0.12
Activations Density 0.242%