INDEX

Explanations

winning or finding something

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

how

-1.46

וֹ

-1.45

its

-1.33

now

-1.32

the

-1.31

oires

-1.27

 your

-1.24

 césped

-1.20

וּ

-1.20

 السياس

-1.18

POSITIVE LOGITS

 fufficient

1.36

 registró

1.26

 сообщили

1.23

 ویتامین

1.16

 성공

1.16

 nonstop

1.15

 homoto

1.15

ڑ

1.14

 occafion

1.13

"}";

1.13

Activations Density 0.017%