INDEX
Explanations
words related to ice, cold environments, and frozen substances
New Auto-Interp
Negative Logits
ributed
-0.77
enance
-0.74
ours
-0.72
ysis
-0.71
###
-0.70
ible
-0.70
tremend
-0.68
ients
-0.68
ivating
-0.68
icals
-0.67
POSITIVE LOGITS
rink
1.08
cream
0.87
ipeg
0.83
Cream
0.82
cream
0.81
flake
0.78
Hockey
0.78
gren
0.78
Blackhawks
0.78
skating
0.77
Activations Density 2.571%