INDEX

Explanations

for + descriptor

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

-1.49

 vicente

-1.47

 different

-1.40

 emailAlready

-1.38

 comprendere

-1.31

 officio

-1.29

 различные

-1.25

 edelstahl

-1.23

 studia

-1.23

 através

-1.22

POSITIVE LOGITS

스

1.37

 Initially

1.33

 Europy

1.20

२

1.19

 accomplishing

1.18

つの

1.17

.}(

1.16

 this

1.13

to

1.13

 شروع

1.10

Activations Density 0.033%