INDEX
Explanations
words related to being cornered or trapped
terms related to legal or formal proceedings
New Auto-Interp
Negative Logits
heit
-0.71
Sov
-0.70
bol
-0.69
grain
-0.69
gebra
-0.67
kies
-0.67
indebted
-0.67
oba
-0.67
velt
-0.65
obyl
-0.65
POSITIVE LOGITS
nered
2.41
SER
2.33
RO
1.99
PRO
1.60
SER
1.40
CTR
1.27
nic
1.04
CONT
0.97
SEM
0.95
Yo
0.94
Activations Density 0.036%