INDEX
Explanations
mountains
This neuron fires on mentions of mountains—the word “mountain” itself or the names of specific peaks.
New Auto-Interp
Negative Logits
Section
-0.06
Volunteer
-0.06
Authority
-0.06
irrig
-0.06
ی
-0.06
Ek
-0.06
territories
-0.06
.material
-0.06
controlling
-0.06
/github
-0.06
POSITIVE LOGITS
آلة
0.07
_take
0.06
incl
0.06
แพร
0.06
racuse
0.06
ickname
0.06
October
0.06
частина
0.06
圭圭
0.06
・・・
0.06
Activations Density 0.014%