INDEX

Explanations

brand names and products

The neuron fires on specific commercial brand or product names.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 welkom

-1.39

豨

-1.37

 there

-1.28

ље

-1.26

 extraordinaire

-1.26

 illustratie

-1.20

 they

-1.18

</h1>

-1.18

 szeret

-1.16

ayas

-1.15

POSITIVE LOGITS

を

1.44

日から

1.39

with

1.34

Or

1.29

以及

1.24

 wartości

1.24

altro

1.20

 Wartawan

1.17

®,

1.16

을

1.16

Activations Density 0.048%