INDEX
Explanations
phrases related to ants and their behaviors
New Auto-Interp
Negative Logits
ankan
-0.16
fencing
-0.16
duck
-0.15
usb
-0.14
usb
-0.13
mus
-0.13
Gem
-0.13
duck
-0.13
libertine
-0.13
ubo
-0.13
POSITIVE LOGITS
ants
0.35
queen
0.30
worker
0.29
Worker
0.29
queen
0.29
colony
0.28
Colony
0.27
workers
0.26
worker
0.26
queens
0.25
Activations Density 0.034%