INDEX

Explanations

retail and fashion publications

The neuron specifically activates on capitalized proper names of retailers, brands, and publications.

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

-1.77

-1.66

 Furthermore

-1.52

 However

-1.47

-1.42

-1.41

also

-1.40

-1.39

 Also

-1.38

POSITIVE LOGITS

Bises

1.42



1.40

Comics

1.33

ENR

1.32

zysta

1.29

 Nordstrom

1.29

кзак

1.26

 péché

1.23

 denken

1.22

denim

1.21

Activations Density 0.014%