INDEX

Explanations

GPT-4

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 veggies

1.37

 Transportation

1.29

蔬菜

1.29

 пище

1.26

 Wellness

1.24

 Floral

1.23

 фото

1.23

 moms

1.21

赤ちゃん

1.21

 газо

1.21

POSITIVE LOGITS

 insofar

1.28

 intellectually

1.21

 epistem

1.12

 undertakes

1.12

 journalistic

1.10

 ontological

1.09

 analytically

1.09

 propagand

1.09

>(&

1.07

 legitimately

1.07

Activations Density 0.658%