INDEX

Explanations

Of these, more than

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

-1.51

ramientas

-1.31

 again

-1.24

有任何

-1.23

 Vương

-1.23

Müller

-1.21

thren

-1.20

 ANYTHING

-1.18

 تمامی

-1.18

 OCTOBER

-1.16

POSITIVE LOGITS

use

1.40

to

1.39

 other

1.35

Other

1.34

 takich

1.28

There

1.23

 otras

1.20

 diantaranya

1.20

 have

1.18

 various

1.16

Activations Density 0.084%