INDEX
Explanations
The neuron is triggered by mentions of the material “paper” (and close variants like “paperboard”).
New Auto-Interp
Negative Logits
肌
-0.06
损
-0.06
undefeated
-0.06
详情
-0.06
uninstall
-0.06
-flight
-0.06
964
-0.06
Tx
-0.06
GraphQL
-0.06
Holmes
-0.06
POSITIVE LOGITS
paper
0.13
Paper
0.12
papel
0.09
-paper
0.08
Paper
0.08
纸
0.08
paper
0.08
紙
0.08
answers
0.07
apyrus
0.07
Activations Density 0.010%