INDEX
Negative Logits
anisotropic
0.37
disamb
0.37
berücksichtigt
0.37
'='
0.36
negated
0.36
divergences
0.35
ReLU
0.34
misleading
0.34
subsum
0.33
delimiters
0.33
POSITIVE LOGITS
avevano
0.43
aveva
0.42
brook
0.40
సంవత్సర
0.40
കുടുംബ
0.38
courtyard
0.38
साल
0.37
ôtel
0.36
生活
0.36
પરિવાર
0.36
Activations Density 0.451%