INDEX

Explanations

positive impact or effects

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

-2.08

or

-1.92

but

-1.45

 unterschiedlich

-1.42

擄

-1.41

 verschieden

-1.40

 geweldige

-1.38

we

-1.38

the

-1.35

POSITIVE LOGITS

褲

1.74

الدرا

1.60

騭

1.55

Traducción

1.48

𐑂

1.47

tovat

1.46

المشار

1.44

蒾

1.43

Añ

1.42

أي

1.42

Activations Density 0.050%