INDEX

Explanations

first followed by a word or punctuation

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

æŃ¤æĹ¶

-0.27

è¿ĻæĹ¶

-0.26

åĲĮæĹ¶ä¹Ł

-0.24

æľĢç»Ī

-0.24

 materially

-0.24

ç±»ä¼¼

-0.24

ä¸Ńæĸ°

-0.23

 otherwise

-0.23

åĲĮæĹ¶

-0.23

 meanwhile

-0.23

POSITIVE LOGITS

åħĪæĺ¯

0.31

åħĪæĬĬ

0.28

è¦ģçŁ¥éģĵ

0.27

åħĪ

0.27

 ë¨¼ìłĢ

0.26

ç¬¬ä¸ĢæŃ¥

0.25

 Firstly

0.25

teen

0.24

é¦ĸåħĪæĺ¯

0.23

ë¶ĢíĦ°

0.23

Activations Density 0.049%