INDEX

Explanations

locations and code

opening parentheses in structured data contexts like lists, citations, and code.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

-2.19

-1.52

-1.50

-1.39

’

-1.30

for

-1.30

$^{

-1.29

-1.28

нибудь

-1.23

św

-1.23

POSITIVE LOGITS

 patienter

1.65

ⓓ

1.57

 مختلف

1.53

躥

1.52

Mitar

1.47

bilden

1.42

焢

1.42

这些

1.42

他们

1.41

 expéri

1.40

Activations Density 0.073%