INDEX
Explanations
references to drone usage and authority in the context of national security
New Auto-Interp
Negative Logits
ctor
-0.15
peak
-0.15
azer
-0.14
iverz
-0.14
Detector
-0.14
Peak
-0.14
ems
-0.14
normal
-0.14
airy
-0.14
Injector
-0.14
POSITIVE LOGITS
-target
0.21
targets
0.21
target
0.21
targeting
0.20
/target
0.20
targets
0.20
Targets
0.19
Targets
0.18
Target
0.18
Target
0.18
Activations Density 0.026%