INDEX

Explanations

paying attention or listening

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

of

-1.59

 everything

-1.36

 kemudian

-1.34

how

-1.32

ハロウィン

-1.30

 Stoppers

-1.29

⽉

-1.27

 propiet

-1.25

流石

-1.23

POSITIVE LOGITS

liğ

1.38

鮠

1.33

solete

1.29

 Gesetzes

1.26

ем

1.25

 ещё

1.25

The

1.24

 piedi

1.23

Еще

1.23

 Gunakan

1.23

Activations Density 0.014%