INDEX
Explanations
Period symbols
This neuron doesn’t respond to any content—it remains inactive and doesn’t fire for any tokens.
New Auto-Interp
Negative Logits
(Un
-0.07
innings
-0.07
freight
-0.06
ntag
-0.06
предвар
-0.06
paddingBottom
-0.06
fre
-0.06
паци
-0.06
/Auth
-0.06
Proc
-0.06
POSITIVE LOGITS
fwrite
0.07
:",
0.07
modifiers
0.06
overwritten
0.06
.collider
0.06
exemple
0.06
multiply
0.06
ξύ
0.06
gement
0.06
verify
0.06
Activations Density 0.017%