INDEX

Explanations

for which

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¯¹æŃ¤

-0.12

ä¸ºäºĨè®©

-0.09

åĲĪ

-0.08

æŃ¥æŃ¥

-0.08

/MIT

-0.08

zs

-0.08

å¸ĮæľĽèĩªå·±

-0.08

åľ¨è¿Ļæĸ¹éĿ¢

-0.08

×Ļ×ľ

-0.08

å°±è§īå¾Ĺ

-0.08

POSITIVE LOGITS

 sake

0.15

 purposes

0.14

æĿ¥è¯´

0.13

ä¾Ĩèªª

0.11

æĿ¥è®²

0.10

çİ©å®¶æĿ¥è¯´

0.10

åĩºè·¯

0.10

 Ð¿ÑĢÐµÐ´Ð½Ð°Ð·

0.09

 ÙĪÙĦ

0.09

å¼Ģåĩº

0.09

Activations Density 0.127%