INDEX

Explanations

country names and programming identifiers

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

 then

-2.36

so

-1.71

 before

-1.69

 would

-1.62

であれば

-1.40

 dann

-1.34

那麼

-1.24

 тогда

-1.22

 قبل

-1.17

 entonces

-1.16

POSITIVE LOGITS

即使

1.32

 while

1.32

 this

1.27

 because

1.26

how

1.23

 even

1.23

 will

1.20

 навіть

1.18

but

1.17

 mentre

1.16

Activations Density 0.023%