INDEX

Explanations

many, many people, some language

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

 hereby

-0.26

ä¾Ľå¤§å®¶

-0.25

-0.24

äº¤ç»Ļ

-0.24

è®©äºº

-0.24

ãģ«è¡Į

-0.23

 again

-0.23

å¤§å®¶ä¸Ģèµ·

-0.23

 charges

-0.22

ä½¿äºº

-0.22

POSITIVE LOGITS

èĢĮè¨Ģ

0.49

æĿ¥è¯´

0.46

çİ©å®¶æĿ¥è¯´

0.41

çļĦäººæĿ¥è¯´

0.39

æĪĺå£«æĿ¥è¯´

0.38

æĿ¥è®²

0.36

ä¾Ĩèªª

0.36

 sake

0.34

ãģ«ãģ¨ãģ£ãģ¦ãģ¯

0.32

çľĭæĿ¥

0.29

Activations Density 0.035%