INDEX

Explanations

singular ability or condition

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

 certainly

-1.71

}^{

-1.61

 gewiß

-1.41

している

-1.41

 ervoor

-1.36

蠖

-1.35

哧

-1.35

 chances

-1.33

 در

-1.32

在

-1.30

POSITIVE LOGITS

焢

1.67

谂

1.38

哢

1.28

two

1.27

ophageal

1.27

each

1.25

 furono

1.24

 pères

1.23

两者

1.23

三個

1.23

Activations Density 0.050%