INDEX

Explanations

introducing a resulting action or outcome

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

 that

-2.03

of

-1.95

are

-1.87

 about

-1.84

、“

-1.61

 these

-1.60

 tienen

-1.57

에는

-1.55

<td>

-1.53

 want

-1.52

POSITIVE LOGITS

ėte

1.77

1.72

它

1.63

muerzo

1.61

を注文

1.60

 which

1.59

轳

1.59

</b>

1.53

 우리

1.52

 suggesting

1.49

Activations Density 0.018%