INDEX
Explanations
ignore known, replace weaker/awkward
New Auto-Interp
Negative Logits
Emotion
0.45
condition
0.42
ró
0.40
dis
0.39
सुविधाएं
0.39
সম্পদ
0.38
COMDAT
0.38
Root
0.38
ifts
0.38
important
0.38
POSITIVE LOGITS
aveva
0.56
avevano
0.55
mancanza
0.52
habían
0.52
havia
0.50
avevo
0.49
había
0.47
apunta
0.46
tinha
0.45
嚆
0.44
Activations Density 0.004%