INDEX

Explanations

ten followed by numbers or quantities

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

æ½

-0.08

ÑĢ

-0.08

çĵ´

-0.08

angel

-0.08

 Canc

-0.07

sic

-0.07

duct

-0.07

éĨĴäºĨ

-0.07

Úº

-0.07

 Ð±Ñĭ

-0.07

POSITIVE LOGITS

ãĥĮ

0.09

å¸®åĬ©ä¼ģä¸ļ

0.09

åı¨

0.09

 ÑģÐ¾Ð±Ð¾Ð¹

0.09

iques

0.09

ities

0.09

mpp

0.09

Ð´ÐµÑĢ

0.09

alf

0.08

icity

0.08

Activations Density 0.038%