INDEX

Explanations

auxiliary verbs indicating states

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

 tempel

-1.35

 geograf

-1.32

 variabel

-1.28

läs

-1.28

我们

-1.27

 kompet

-1.25

 foton

-1.24

 artesanato

-1.23

näm

-1.22

 flexibel

-1.21

POSITIVE LOGITS

to

1.43

 since

1.33

 during

1.23

 […]

1.21

 after

1.19

 will

1.16

if

1.16

 when

1.16

You

1.10

 while

1.10

Activations Density 0.095%