INDEX

Explanations

fabricated stories and lies

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Plata

-0.82

losa

-0.81

ôn

-0.79

 conductive

-0.77

ündigen

-0.77

 Fintech

-0.75

 hypocrisy

-0.73

上方

-0.73

 secre

-0.71

꿉

-0.71

POSITIVE LOGITS

 Iraq

0.93

 Steele

0.93

灵

0.90

 dossier

0.88

 story

0.85

 memo

0.84

Taw

0.83

 relies

0.79

 Fabric

0.78

 relied

0.77

Activations Density 0.048%