INDEX

Explanations

modal verb + to/be + consequence

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

 حتی

-1.43

albeit

-1.38

惴

-1.36

 навіть

-1.34

tych

-1.34

骉

-1.32

ㅤ

-1.28

 spektak

-1.27

そして

-1.27

Теперь

-1.26

POSITIVE LOGITS

 other

1.77

it

1.60

one

1.52

 each

1.51

 would

1.42

 from

1.38

 further

1.32

 could

1.27

or

1.27

has

1.27

Activations Density 0.152%