INDEX

Explanations

pages and volume numbers

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

There

-1.52

 there

-1.40

in

-1.34

While

-1.20

 that

-1.18

or

-1.13

 while

-1.13

 dAtA

-1.13

 нарушения

-1.12

 areas

-1.11

POSITIVE LOGITS

}-\

1.27

1.21

 différentes

1.20

 verschillende

1.18

ntp

1.17

 herunter

1.16

 altamente

1.15

1.13

 réun

1.12

 archite

1.12

Activations Density 0.025%