INDEX
Explanations
words related to forward progress, movement, and conclusion
New Auto-Interp
Negative Logits
get
-1.46
create
-1.36
make
-1.34
keep
-1.27
take
-1.26
avoid
-1.23
develop
-1.23
find
-1.22
increase
-1.21
give
-1.20
POSITIVE LOGITS
mourut
1.34
étoit
1.17
venait
1.12
définiti
1.01
ragazze
0.97
auroit
0.96
giudi
0.95
feroit
0.94
avoient
0.93
conmigo
0.93
Activations Density 7.769%