INDEX

Explanations

technical terms or non-english words

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 exemplifies

1.87

 incorporates

1.67

 aspires

1.65

 integrates

1.60

дне

1.60

 embodies

1.57

 সমাধিতে

1.56

 underlies

1.55

 explores

1.55

牘

1.54

POSITIVE LOGITS

 stuff

1.54

!!!!!

1.53

!!!

1.51

!!!!

1.49

-_-

1.40

guy

1.39

!!!!!!!!

1.37

&$

1.36

 tocc

1.34

!!

1.34

Activations Density 0.134%