INDEX

Explanations

finding the balance or middle

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

[{

-1.47

 spannend

-1.46

켭

-1.38

 heridos

-1.30

 larges

-1.30

🄾

-1.27

…»

-1.24

as

-1.23

acknow

-1.23

 عملية

-1.23

POSITIVE LOGITS

 between

2.09

between

1.88

 ktoś

1.48

硨

1.31

 будь

1.30

1.29

Kesimpulan

1.22

なんでも

1.21

OrNil

1.20

Activations Density 0.028%