INDEX

Explanations

congratulations or persistent conditions

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

of

-2.16

 الأشخاص

-1.43

 معظم

-1.38

 considérons

-1.36

kenal

-1.30

釀

-1.27

 Rumuni

-1.27

 क्या

-1.24

 natuurlijk

-1.23

なのに

-1.21

POSITIVE LOGITS

 tarifas

1.44

有趣

1.43

…」

1.38

 sedia

1.38

)";

1.37

↵

1.34

!」

1.34

Да

1.34

在一起

1.31

否定

1.29

Activations Density 0.035%