INDEX
Explanations
emotional expressions and sentiments associated with love and compassion
New Auto-Interp
Negative Logits
ku
-0.17
à¥Ĥस
-0.17
aber
-0.16
elow
-0.15
Dismiss
-0.14
uby
-0.14
üstü
-0.14
æij©
-0.14
ares
-0.14
emaker
-0.14
POSITIVE LOGITS
wrench
0.27
rending
0.26
ache
0.22
felt
0.22
warming
0.21
rend
0.21
break
0.20
broken
0.19
ACHE
0.19
rend
0.19
Activations Density 0.011%