INDEX

Explanations

clothing and accessories

The neuron fires on specific product/item names (e.g., “Ring,” “Boots,” “Tote,” “Cardigan,” etc.) in a shopping/catalog context.

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

 brachten

-1.51

iibo

-1.51

 émissions

-1.45

 preponder

-1.45

pecabe

-1.42

 半袖

-1.41

 hebt

-1.40

 nejlepší

-1.37

zeera

-1.34

 leçons

-1.34

POSITIVE LOGITS

two

1.72

bü

1.59

 using

1.55

 even

1.51

↵↵

1.48

 also

1.47

my

1.43

 still

1.39

 with

1.38

by

1.38

Activations Density 0.062%