INDEX

Explanations

Eureka! Q: EPA

HTML or web-related code elements.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

and

-1.41

its

-1.38

 Each

-1.31

 their

-1.18

LEBR

-1.17

 durch

-1.13

ICIÓN

-1.12

の方も

-1.11

気味

-1.10

 Other

-1.09

POSITIVE LOGITS

excellents

1.37

škoda

1.23

VERY

1.17

 áng

1.17

极为

1.16

但是在

1.16

somebody

1.16

がありますが

1.15

 heridos

1.15

 Kitch

1.15

Activations Density 0.000%