INDEX

Explanations

for or in followed by females or the

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ä¸įä»ħåı¯ä»¥

-0.13

ä¸įä»ħè¦ģ

-0.10

å°±å¥½åĥı

-0.09

ä¹Łè®©

-0.09

ä¸įä»ħèĥ½

-0.09

 nÃ¤mlich

-0.09

è¿Ļä¸ĢçĤ¹

-0.09

è¿ĻäºĽéĹ®é¢ĺ

-0.09

(Operation

-0.08

æĹ¢èĥ½

-0.08

POSITIVE LOGITS

 ones

0.15

 others

0.13

åħ¶ä»ĸçļĦ

0.11

åħ¶ä»ĸäºº

0.11

others

0.10

ones

0.10

 Ones

0.10

ãģĿãģĨãģ§

0.10

 Others

0.08

 Macros

0.08

Activations Density 0.169%

No Known Activations