INDEX

Explanations

PET scan

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æĪĳå¿ĥ

-0.07

åĬĽå£«

-0.06

 Elliott

-0.06

 crucial

-0.06

(always

-0.06

ä¹ħ

-0.06

ç¥ŀçİĭ

-0.06

gue

-0.05

new

-0.05

 twice

-0.05

POSITIVE LOGITS

è½¬è½½æĪĸ

0.08

ç¨³

0.08

nhof

0.08

/***

0.08

çĽ¸åħ³æĸ°éĹ»

0.08

è½¬è½½æĪĸå¼ķçĶ¨

0.07

ç¨³ä½ı

0.07

æĹłæĦ§

0.07

illow

0.07

brane

0.07

Activations Density 0.067%