INDEX
Explanations
time references, particularly in relation to duration and frequency
New Auto-Interp
Negative Logits
Local
-0.66
ople
-0.65
anooga
-0.61
rote
-0.60
Hels
-0.59
liction
-0.58
icial
-0.58
Regional
-0.57
acted
-0.56
claimed
-0.56
POSITIVE LOGITS
ago
1.39
olds
1.12
Ago
1.07
hindsight
0.92
ago
0.88
elapsed
0.85
consecut
0.78
ruary
0.75
yrs
0.75
kids
0.75
Activations Density 0.094%