INDEX
Explanations
The neuron responds to numeric-statistic tokens, i.e. numbers especially with decimal points or percentage‐style figures.
New Auto-Interp
Negative Logits
.Nil
-0.08
كار
-0.07
Plays
-0.07
()"↵
-0.07
_ROW
-0.07
frightened
-0.07
Ips
-0.06
insertion
-0.06
_HERSHEY
-0.06
آزمایش
-0.06
POSITIVE LOGITS
elimination
0.07
anticipated
0.06
Dayton
0.06
lookahead
0.06
cleanup
0.06
кти
0.06
Fac
0.06
_age
0.06
thân
0.06
važ
0.06
Activations Density 0.433%