INDEX

Explanations

varied conversational text snippets

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

are

-0.63

is

-0.63

ktır

-0.52

Says

-0.52

cannot

-0.52

 aren

-0.52

seems

-0.52

 cannot

-0.51

ಿದೆ

-0.51

￡

-0.51

POSITIVE LOGITS

was

1.46

had

1.30

 lacked

1.16

 wanted

1.12

 knew

1.10

 sounded

1.10

 were

1.09

 depended

1.09

 felt

1.09

was

1.08

Activations Density 0.049%