INDEX

Explanations

enhancement, improvement, addition

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

 even

-1.66

we

-1.44

and

-1.42

 what

-1.41

 being

-1.39

as

-1.36

 Even

-1.34

but

-1.31

Not

-1.31

 Being

-1.26

POSITIVE LOGITS

 quaisquer

1.43

 disminuir

1.36

lieving

1.35

笵

1.35

 facilitar

1.34

 weihnachten

1.33

狴

1.31

可能な

1.31

 harten

1.30

 keinerlei

1.29

Activations Density 0.120%