INDEX

Explanations

inquiring about correctness or suitability

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

They

-1.33

 srebr

-1.28

 préf

-1.21

 invece

-1.19

 защото

-1.15

liegt

-1.14

 ถ้า

-1.14

 zás

-1.12

 rám

-1.11

психо

-1.11

POSITIVE LOGITS

 then

1.73

of

1.41

 this

1.30

now

1.28

おはようございます

1.25

Ẫ

1.19

 vair

1.18

ERTY

1.18

 various

1.17

 throughout

1.16

Activations Density 0.113%