INDEX

Explanations

social media platforms and websites

The neuron flags marketing call-to-action phrases (e.g. “follow on Twitter,” “purchase at …,” “available at …”).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 разрабо

0.66

]}$.

0.60

修

0.59

 הראש

0.57

Rounded

0.57

 народов

0.56

ודה

0.54

 ganglia

0.54

Rewrite

0.54

 Ganga

0.54

POSITIVE LOGITS

 Facebook

1.36

 Twitter

1.35

 twitter

1.24

facebook

1.23

 facebook

1.20

 LinkedIn

1.17

 YouTube

1.16

Facebook

1.15

 Linkedin

1.14

 Instagram

1.13

Activations Density 0.134%