INDEX

Explanations

mathematical expression equality

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

æĭ¼éŁ³

-0.08

{:

-0.07

éķľæ£ĢæŁ¥

-0.07

apon

-0.07

{@

-0.07

ropol

-0.07

 onCreate

-0.07

æľºæŀĦåĴĮ

-0.06

Ã£n

-0.06

åĪ©

-0.06

POSITIVE LOGITS

 equality

0.08

 numerator

0.08

estar

0.07

éĻĮ

0.07

 expression

0.07

 supposedly

0.07

ighted

0.07

èµıèĬ±

0.07

 value

0.07

 sums

0.07

Activations Density 0.047%

No Known Activations