INDEX

Explanations

modal verbs followed by action verbs

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

åĪ°åľº

-0.23

é¢ĨåŁŁçļĦ

-0.22

äººçĶŁçļĦ

-0.22

åĽ½å®¶éĺŁ

-0.22

å¤§å®¶ä¸Ģèµ·

-0.20

äººåĢĳ

-0.20

åħļåĴĮ

-0.20

ä¸ªäººè§ĤçĤ¹

-0.20

ä¼ģä¸ļåĴĮ

-0.20

ä¸ĭéĿ¢æĺ¯

-0.20

POSITIVE LOGITS

èĩªèº«

0.27

 itself

0.25

 ÐĲÐ»ÐµÐºÑģÐ°Ð½

0.25

ä¸ºåħ¶

0.24

å¯¹åħ¶

0.23

 ìŀĲì²´

0.22

 ÐĶÐ¼Ð¸ÑĤÑĢ

0.22

à¹Ģà¸Ĺà¸Ħà¹Ĥà¸Ļ

0.21

SCRI

0.21

åľ¨åħ¶

0.20

Activations Density 0.148%