INDEX

Explanations

receivers of communication

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

åĮĸè¿Ľç¨ĭ

-0.23

æ¦Ĥå¿µ

-0.22

xac

-0.22

çĪ±åĽ½ä¸»ä¹ī

-0.22

ä¸»ä¹īæĢĿæĥ³

-0.22

mente

-0.21

èº«ä»·

-0.21

à¸ªà¸£à¸£

-0.21

åı»

-0.21

å¼ĢæĪĺ

-0.21

POSITIVE LOGITS

è¿ľæĸ¹

0.31

ä»£è¡¨

0.27

{o

0.25

 people

0.24

å®¢æĪ·

0.24

 colleagues

0.24

 angry

0.24

 representative

0.24

 clients

0.23

onen

0.23

Activations Density 0.035%