INDEX

Explanations

indefinite article usage

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

 übri

-1.58

 gesamte

-1.54



-1.45

 incrí

-1.41

 other

-1.36

 Pristup

-1.33

 redire

-1.33

 eigentliche

-1.31

 what

-1.30

 nutric

-1.30

POSITIVE LOGITS

 использование

1.31

1.27

the

1.26

The

1.24

 ਨੂੰ

1.23

nullptr

1.23

穣

1.22

adple

1.21

 подготовка

1.20

 ($("#

1.20

Activations Density 0.358%