examples, comparisons, and explanations

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

erti

-0.07

Î®ÏĦÎ·

-0.07

.decorators

-0.07

 ÑĤÐµÐ¼Ð°

-0.07

 Topic

-0.07

Ã¨se

-0.07

sic

-0.07

ipes

-0.07

éĢļè¿ĩåĲĦç§į

-0.07

POSITIVE LOGITS

ä¹Łä¸įå¤ļ

0.07

 examples

0.07

ÑħÐ¾Ð´Ð¾Ð²

0.07

 rationale

0.07

 explanation

0.06

à¥ĭà¤°

0.06

 explanations

0.06

 comparisons

0.06

å¦¥

0.06

à¸ĺà¸´

0.06

Activations Density 0.020%

No Known Activations