INDEX

Explanations

perceived threat information

perceived stimuli or cues

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 commandant

0.46

 שלא

0.44

genden

0.44

 totaal

0.43

спеди

0.42

 setzen

0.41

 POLICY

0.41

탄

0.41

 alla

0.40

 تمامی

0.40

POSITIVE LOGITS

 ambiguous

0.57

 salient

0.55

 напом

0.53

 perceived

0.52

 ambiguity

0.52

 repulsive

0.52

 percib

0.52

 атрибу

0.50

 cues

0.48

 stimuli

0.48

Activations Density 0.139%