INDEX
Explanations
scientific/medical papers
This neuron fires on specialized scientific or technical nouns denoting fields, measurements, or large-scale studies (e.g. terms ending in suffixes like ‑ome, ‑metry, ‑ology, or ‑kines).
New Auto-Interp
Negative Logits
(("-0.06
postgres
-0.06
racial
-0.06
↵↵
-0.06
boards
-0.06
''' ↵
-0.06
.*/↵
-0.06
belts
-0.06
)&
-0.06
社会
-0.06
POSITIVE LOGITS
Disqus
0.07
éry
0.07
ò
0.07
onic
0.07
επί
0.07
inium
0.07
อง
0.07
алов
0.07
ộ
0.07
chương
0.07
Activations Density 0.206%