INDEX

Explanations

internet domain names

The neuron detects fragments of web and email addresses—particularly domain name components (e.g. “.com,” “.org,” hostnames and TLDs).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 livestream

-0.85

clyde

-0.82

‍♂️

-0.81

 kush

-0.80

 sooo

-0.79

вики

-0.79

 наших

-0.78

 soooo

-0.77

teva

-0.77

 indonesia

-0.75

POSITIVE LOGITS

net

1.05

net

1.04

org

0.98

putchar

0.92

com

0.91

Berdasarkan

0.86

 kolorze

0.85

wishlist

0.85

 росій

0.85

zdroj

0.83

Activations Density 0.090%