INDEX

Explanations

flipping through magazine stories

This neuron responds to trademarked brand, product, or company names.

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

陞

-1.41

漉

-1.38

 aloj

-1.32

 entweder

-1.29

it

-1.27

淦

-1.25

 mengatasi

-1.24

忸

-1.24

noj

-1.23

良好的

-1.23

POSITIVE LOGITS

 มา

1.41

 hacemos

1.34

 första

1.30

 drit

1.28

 ثلاثة

1.20

私

1.19

 vinci

1.14

亲

1.11

粜

1.10

偶尔

1.09

Activations Density 0.087%