INDEX
Explanations
network protocols and termsormedical terms and entities
proper names (including surnames, technical terms, and product names).
New Auto-Interp
Negative Logits
雰
-2.86
硖
-2.83
鈇
-2.73
ſame
-2.70
潆
-2.69
蛲
-2.66
戗
-2.63
تیجه
-2.53
俢
-2.52
洳
-2.52
POSITIVE LOGITS
).
2.88
<i>
2.73
ra
2.69
↵
2.64
m
2.59
i
2.58
u
2.58
b
2.56
on
2.55
wysokiej
2.48
Activations Density 0.047%