INDEX

Explanations

phrases related to the production and responsibilities of news programs

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

rather

-0.51

 rather

-0.50

 instead

-0.49

 increasingly

-0.49

instead

-0.43

 almost

-0.41

 piuttosto

-0.41

HasAnnotation

-0.41

 plutôt

-0.41

 кое

-0.39

POSITIVE LOGITS

 quelconque

0.81

any

0.77

 anything

0.75

 alcun

0.72

 quaisquer

0.72

 siquiera

0.71

 alcuna

0.70

 anymore

0.68

 mitään

0.68

 żad

0.67

Activations Density 4.130%

No Known Activations