INDEX
Explanations
romantic
The neuron fires on Portuguese-language stems of “romantic” (e.g. the “românt-” parts of romântica/romântico), i.e. romance-themed words in Portuguese.
New Auto-Interp
Negative Logits
пе
-0.06
Warn
-0.06
Detach
-0.06
Thr
-0.06
Hatch
-0.06
On
-0.05
ose
-0.05
itto
-0.05
Igor
-0.05
!".
-0.05
POSITIVE LOGITS
đâu
0.07
moved
0.07
цент
0.06
pronounced
0.06
ёт
0.06
Explicit
0.06
★
0.06
adge
0.06
-party
0.06
Builder
0.06
Activations Density 0.019%