INDEX

Explanations

reporting "according to"

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

 because

-1.63

It

-1.56

 perquè

-1.52

</b>

-1.41

巳

-1.40

and

-1.34

 porque

-1.34

</h1>

-1.32

so

-1.28

now

-1.23

POSITIVE LOGITS

 alemania

1.47

Waktu

1.45

するという

1.35

 that

1.32

Karakteristik

1.30

゛

1.28

 angepasst

1.25

によると

1.25

Visión

1.25

個人的には

1.23

Activations Density 0.071%