INDEX

Explanations

bright colors and their descriptions

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Gray

-0.12

 Stern

-0.12

 dusty

-0.11

 Stevenson

-0.11

fog

-0.10

 Franco

-0.10

vag

-0.10

dim

-0.10

 solemn

-0.10

 Gray

-0.10

POSITIVE LOGITS

-red

0.13

 orange

0.13

-white

0.12

-orange

0.11

-blue

0.11

-green

0.11

 green

0.11

red

0.11

 yellow

0.10

 blue

0.10

Activations Density 0.015%