INDEX

Explanations

specific terms or abbreviations

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

it

-1.41

of

-1.34

ּוֹ

-1.33

-1.27

ContentAsync

-1.26

for

-1.25

}$,

-1.23

 бампер

-1.23

 "..\..\..\

-1.20

 бел

-1.20

POSITIVE LOGITS

Curios

1.38

𝑈

1.37

】：

1.32

getreten

1.32

 bluse

1.27

σεις

1.27

 partiellement

1.24

ρες

1.23

Джерела

1.22

《

1.22

Activations Density 0.003%