INDEX
Explanations
story choices or prose perfection
New Auto-Interp
Negative Logits
Kaspersky
0.50
භාවිත
0.49
quiet
0.48
заседа
0.48
quiet
0.47
thuê
0.46
treball
0.46
militare
0.46
tratar
0.45
nevoie
0.45
POSITIVE LOGITS
Spr
0.47
Spr
0.45
browns
0.43
ញ
0.42
ajuan
0.42
(
0.41
Control
0.41
Test
0.41
Spray
0.41
dal
0.41
Activations Density 0.005%