INDEX

Explanations

action verbs followed by common words

The neuron responds to present-tense action verbs that describe adding or including code or configuration (e.g. “include,” “add,” “copy,” “use,” “compile”).

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

not

-1.55

那些

-1.41

 برای

-1.39

有些

-1.34

 questi

-1.34

!]

-1.32

-1.28

之后

-1.25

</tbody>

-1.23

最好的

-1.23

POSITIVE LOGITS

of

1.44

РЫ

1.41

れてる

1.38

ucksack

1.36

it

1.35

 WALTZ

1.34

せば

1.32

тся

1.30

pernic

1.28

🎷

1.28

Activations Density 1.315%