INDEX
Explanations
named entities or terms consisting of the word "arch."
references to architectural terms or concepts
New Auto-Interp
Negative Logits
lling
-0.90
ting
-0.77
nder
-0.76
terness
-0.76
lli
-0.74
instein
-0.71
Tube
-0.71
cers
-0.70
enez
-0.70
ensing
-0.70
POSITIVE LOGITS
itect
1.55
ipel
1.08
ief
0.89
ocobo
0.86
osaurs
0.85
arch
0.81
angel
0.77
opoulos
0.76
isol
0.75
Corp
0.74
Activations Density 0.023%