INDEX

Explanations

suggest, indicate, highlight

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

 along

-1.40

his

-1.37

 kwenye

-1.34

 already

-1.32

 across

-1.30

の場合

-1.30

 inside

-1.27

of

-1.24

 genauso

-1.23

之以

-1.23

POSITIVE LOGITS

一種

1.48

Tür

1.44

Aktual

1.43

してみてください

1.43

Jadi

1.39

Teraz

1.38

Bagi

1.38

Penulis

1.37

 作り方

1.37

Metode

1.36

Activations Density 0.064%