INDEX

Explanations

how many

The neuron is triggered by words that frame or refer to quantities and measurements (e.g. “how,” “many,” “count,” “long,” “times”)—i.e. tokens introducing or describing numeric/quantitative concepts.

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

𝘣

-1.41

 slova

-1.36

諫

-1.35

 cómoda

-1.32

號

-1.29

撰文

-1.27

 شکست

-1.25

 Bagaimana

-1.25

gabung

-1.23

まずは

-1.23

POSITIVE LOGITS

 when

1.31

1.27

 Forder

1.25

ly

1.25

its

1.23

ртқы

1.19

急忙

1.19

δες

1.18

 were

1.18

ore

1.18

Activations Density 0.167%