INDEX

Explanations

abstract concepts and definitions

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

也非常

-1.39

 libéral

-1.33

 infecciones

-1.31

 känner

-1.27

 inflama

-1.25

Schwer

-1.24

 steadfast

-1.23

続きます

-1.22

 didn

-1.21

 thwarted

-1.21

POSITIVE LOGITS

 which

2.81

of

2.80

 that

2.36

or

1.95

 where

1.70

1.58

taya

1.55

tya

1.48

1.44

rrrrrr

1.38

Activations Density 0.100%