INDEX
Explanations
references to age, tradition, and nostalgia
New Auto-Interp
Negative Logits
older
-0.24
older
-0.24
Older
-0.19
oldest
-0.19
ively
-0.16
old
-0.16
ably
-0.16
urally
-0.15
eldre
-0.15
criptor
-0.15
POSITIVE LOGITS
-fashioned
0.54
fashioned
0.47
-school
0.43
-fashion
0.39
school
0.35
timer
0.33
-style
0.33
/new
0.32
sters
0.32
tim
0.30
Activations Density 0.071%