INDEX

Explanations

.com and .org, or numbers

This neuron detects domain tokens in URLs and email addresses (e.g., “.com,” “yahoo,” “org”).

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

 that

-1.50

 cenário

-1.21

 relatório

-1.19

 território

-1.18

 domínio

-1.15

 estúdio

-1.11

 planejamento

-1.11

}^{

-1.09

 treinamento

-1.08

 edifício

-1.08

POSITIVE LOGITS

or

2.53

and

1.58

もしくは

1.32

或是

1.25

 atau

1.23

 vacas

1.23

あるいは

1.22

 शुर

1.20

 cortar

1.17

 یا

1.14

Activations Density 0.002%