INDEX

Explanations

depends on what/if/how/who

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

either

-1.41

Either

-1.39



-1.33

is

-1.27

 malgré

-1.23

噥

-1.23

aucun

-1.23

-1.22

nitus

-1.20

 hilsen

-1.18

POSITIVE LOGITS

 their

1.55

']

1.30

กลับ

1.29

if

1.25

 机械

1.23

ک

1.23

Демографія

1.23

''

1.21

 경우

1.20

 особенно

1.20

Activations Density 0.010%