INDEX

Explanations

direction or axis

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

åĿ¼

-0.09

åĩı

-0.09

CCP

-0.08

ilities

-0.08

 Ð¼Ð¾Ð´

-0.08

éĻĨç»Ń

-0.08

guÃ©

-0.08

æį¢æĪĲ

-0.08

ä¸įçŁ¥

-0.08

ROUT

-0.08

POSITIVE LOGITS

æĸ¹åĲĳ

0.12

 directions

0.11

 direction

0.10

ìĭ¬

0.09

æ·ĭ

0.09

çļĦæĸ¹åĲĳ

0.09

direction

0.08

-direction

0.08

å®¡æī¹

0.08

åĴ¨è¯¢æľįåĬ¡

0.08

Activations Density 0.215%