INDEX
Explanations
This neuron fires primarily on mentions of software libraries or development platforms (e.g., words like “library,” “platform,” or similar).
New Auto-Interp
Negative Logits
OUNTRY
-0.07
Languages
-0.07
noises
-0.07
804
-0.07
Deaths
-0.06
(report
-0.06
tako
-0.06
countries
-0.06
mushrooms
-0.06
ORA
-0.06
POSITIVE LOGITS
0.06
actory
0.06
Nim
0.06
レディース
0.06
)set
0.06
sched
0.06
_On
0.06
гал
0.06
conv
0.06
.des
0.05
Activations Density 0.009%