INDEX
Explanations
elements related to architectural features and furnishings
New Auto-Interp
Negative Logits
houſe
-1.14
myſelf
-1.12
Theſe
-1.11
Efq
-1.11
Monfieur
-1.10
itſelf
-1.09
Houſe
-1.09
Majefty
-1.09
ValueStyle
-1.09
ſelf
-1.07
POSITIVE LOGITS
0.57
T
0.53
in
0.52
and
0.48
V
0.48
&
0.47
/
0.46
<eos>
0.46
all
0.45
/
0.43
Activations Density 0.044%