INDEX

Explanations

connecting phrases with and

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

тування

-1.66

</u>

-1.36

 Quizzes

-1.35

</h3>

-1.32

{?

-1.30

鬭

-1.29

》,

-1.26

，”

-1.25

 wichtiger

-1.25

土曜日

-1.24

POSITIVE LOGITS

of

2.42

1.66

accepte

1.45

"'

1.41

staden

1.37

1.35

".

1.34

 probablemente

1.29

 Für

1.28

خول

1.28

Activations Density 0.079%