INDEX

Explanations

social etiquette and norms

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

æĬ¥è¡¨

-0.27

çļĦçĶŁåĳ½

-0.25

æĤ²è§Ĥ

-0.25

çİ°è´§

-0.25

åıĤä¿Ŀ

-0.25

æĦģ

-0.24

çĶŁåĳ½

-0.24

è·ĳäºĨ

-0.24

çªŁ

-0.24

æĢªçī©

-0.24

POSITIVE LOGITS

 etiquette

0.71

ç¤¼ä»ª

0.70

ç¤¼è²Į

0.48

äº¤éĻħ

0.47

åľºåĲĪ

0.45

ä¸¾æŃ¢

0.45

 norms

0.45

 greeting

0.42

äº¤å¾Ģ

0.41

 polite

0.40

Activations Density 0.080%