INDEX

Explanations

names and associated identifiers

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

it

-1.25

Men

-1.04

 children

-1.01

Be

-0.99

utile

-0.98

⟹

-0.98

undial

-0.96

disciplinar

-0.96

men

-0.95

 permiten

-0.94

POSITIVE LOGITS

Does

1.46

Gets

1.30

Obviously

1.26

Because

1.20

Thats

1.18

Wirk

1.16

Usually

1.13

Antwort

1.13

Generally

1.13

Would

1.11

Activations Density 0.044%