INDEX
Explanations
logo design
The neuron responds to parts of the text where the assistant is presenting or discussing logo design options (e.g. the word “logo” and option/list headings).
New Auto-Interp
Negative Logits
syscall
-0.06
ionální
-0.06
Sink
-0.06
######
-0.06
Port
-0.06
射
-0.06
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
-0.06
ers
-0.06
_worker
-0.06
716
-0.06
POSITIVE LOGITS
lij
0.07
ژاپ
0.06
Τι
0.06
Trojan
0.06
neau
0.06
ialized
0.06
modular
0.06
cke
0.06
مل
0.06
olum
0.06
Activations Density 0.027%