INDEX

Explanations

regex patterns

The neuron fires on bits of regular‐expression syntax—backslashes, character‐class brackets, quantifiers, and other regex metacharacters.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 edifício

0.42

 adaptación

0.42

 associé

0.41



0.40

ín

0.39

 pioneering

0.38

 membangun

0.38

 Zoological

0.37

 प्रार्थ

0.37

POSITIVE LOGITS

ANY

0.54

any

0.48

Delimiter

0.46

ार्ट

0.46

abytes

0.46

regex

0.46

 шаблон

0.46

 کسی

0.45

 اما

0.45

Any

0.45

Activations Density 0.001%