INDEX

Explanations

code structure elements

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

 chré

-1.45

 бого

-1.41

摂取

-1.41

to

-1.30

 стихотво

-1.30

 teolog

-1.27

anskrit

-1.25

by

-1.22

 температуре

-1.18

cznej

-1.15

POSITIVE LOGITS

ﻷ

1.64

🪛

1.40

áles

1.39

 their

1.39

ђа

1.37

 informou

1.33

 almohada

1.32

 Jerusalén

1.29

 damska

1.27

ﻹ

1.26

Activations Density 0.001%