INDEX

Explanations

describing quantity or degree

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

 will

-1.56

 when

-1.45

 with

-1.45

 while

-1.41

of

-1.36

 first

-1.35

 during

-1.34

 other

-1.30

 additional

-1.29

when

-1.24

POSITIVE LOGITS

 uniquement

1.41

 extremely

1.41

relatively

1.41

 relativement

1.38

 apenas

1.34

allows

1.30

 VERY

1.30

 plutôt

1.28

 केवल

1.25

 exceptionally

1.20

Activations Density 0.087%