INDEX
Explanations
various forms of personal or emotive expressions
New Auto-Interp
Negative Logits
Łèĥ½
-0.16
ubu
-0.16
lick
-0.16
ép
-0.16
rello
-0.15
neau
-0.14
arius
-0.14
uyen
-0.14
ark
-0.14
strom
-0.14
POSITIVE LOGITS
LOCAL
0.19
local
0.18
prep
0.16
-produ
0.15
past
0.15
locally
0.15
local
0.15
LOCAL
0.15
Local
0.15
BATCH
0.15
Activations Density 0.006%