INDEX

Explanations

negative consequences and effects

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

å±ŀäºİèĩªå·±

-0.25

å¹´ä¹ĭä¹ħ

-0.24

åĲįåı«

-0.23

éĢĤåĲĪèĩªå·±

-0.23

è¿Īè¿Ľ

-0.22

è§Ħæ¨¡æľĢå¤§

-0.21

äºĨè§£ä¸Ģä¸ĭ

-0.21

åį´è¢«

-0.21

åĽłä¸ºä»ĸ

-0.21

æĹłä¸į

-0.20

POSITIVE LOGITS

å¯¼èĩ´

0.33

å¸¦æĿ¥çļĦ

0.31

ä½¿å¾Ĺ

0.30

çŃīåĽłç´ł

0.28

å¸¦æĿ¥

0.26

ä¼ļå½±åĵį

0.25

ä½¿

0.25

ä¼ļå¯¼èĩ´

0.25

å°İèĩ´

0.25

ä¼ļä½¿

0.24

Activations Density 0.484%