INDEX

Explanations

questions seeking clarification

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pera

-0.06

ongs

-0.06

oller

-0.06

å¼ºåİ¿

-0.06

inate

-0.06

åĴĮæĵįä½ľ

-0.06

 vind

-0.06

icate

-0.06

ervations

-0.06

incess

-0.06

POSITIVE LOGITS

æĿĢæīĭ

0.05

 Whites

0.05

>>)

0.05

åĴ

0.05

åŁ¹

0.05

omin

0.04

å¾Ĵæīĭ

0.04

 ones

0.04

 Dimension

0.04

 unbiased

0.04

Activations Density 0.012%