INDEX
Explanations
dates and locations in news reports
dates and locations mentioned in the text
New Auto-Interp
Negative Logits
Reloaded
-0.72
chairs
-0.67
ourselves
-0.60
76561
-0.59
Sund
-0.58
testers
-0.56
geist
-0.56
yourselves
-0.55
rite
-0.54
whiff
-0.53
POSITIVE LOGITS
REUTERS
0.94
©
0.79
FILE
0.77
FILE
0.76
./
0.72
âĶ
0.71
UTERS
0.69
Courtesy
0.66
Photographer
0.66
window
0.65
Activations Density 0.033%