INDEX
Explanations
This neuron activates on mentions of intravenous administration (e.g. the word “intravenous” and its sub-tokens).
New Auto-Interp
Negative Logits
ont
-0.07
avid
-0.06
Blind
-0.06
propiedad
-0.06
ostí
-0.06
_Client
-0.06
ládání
-0.06
tep
-0.06
pressures
-0.06
uma
-0.06
POSITIVE LOGITS
(rhs
0.06
pwm
0.06
.find
0.06
teaspoons
0.06
минут
0.06
/**/*.
0.06
.release
0.06
)); ↵
0.06
_soup
0.06
.repositories
0.06
Activations Density 0.008%