INDEX

Explanations

"I." followed by capitalized words

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

-1.55

vább

-1.40

 остальных

-1.39

珖

-1.29

 hoje

-1.28

 для

-1.27

komfort

-1.26

錚

-1.24

was

-1.22

 pentru

-1.20

POSITIVE LOGITS

かっ

1.59

 brillant

1.55

 uninten

1.46

さてさて

1.40

 démarrer

1.38

 résister

1.37

 stupe

1.33

 réserver

1.33

 reinst

1.32

 diagno

1.31

Activations Density 0.116%