INDEX

Explanations

software components and libraries

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

兩

-1.58

two

-1.36

Three

-1.27

 three

-1.22

Two

-1.21

 caminar

-1.21

 corregir

-1.20

 hacerle

-1.20

两

-1.19

 fabricar

-1.18

POSITIVE LOGITS

</b>

1.23

 suci

1.20

in

1.20

 Dinamarca

1.16

 tanong

1.16

 definiti

1.15

 pastillas

1.15

 frond

1.13

 (…)

1.13

$<$

1.12

Activations Density 0.326%