INDEX

Explanations

pronoun followed by a verb

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

çĽ®åīį

-0.17

çĽ®åīįæŃ£åľ¨

-0.15

 currently

-0.14

currently

-0.13

çĽ®åīįä¸ºæŃ¢

-0.13

çĽ®åīįå·²

-0.12

æĺİå¹´

-0.12

åĲ¬åĲ¬

-0.11

 Currently

-0.11

Ð´Ð°ÐµÑĤÑģÑı

-0.11

POSITIVE LOGITS

å½ĵæĹ¶çļĦ

0.21

å½ĵæĹ¶

0.20

æĪĳå½ĵæĹ¶

0.18

ãģĹãģ¦ãģĦãģ¾ãģĹãģŁ

0.17

ãģ§ãģĹãģŁ

0.17

was

0.17

çķ¶æĻĤ

0.16

ãģĮãģĤãĤĬãģ¾ãģĹãģŁ

0.15

ãģ¦ãģĦãģ¾ãģĹãģŁ

0.15

ãģłãģ£ãģŁãģ®ãģ§

0.15

Activations Density 1.609%