INDEX

Explanations

a/an + adjective

This neuron detects the indefinite articles “a” and “an.”

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

of

-1.66

玙

-1.48

 взаимодействия

-1.39

‘

-1.29

᯾

-1.23

“

-1.23

<h2>

-1.22

 already

-1.20

POSITIVE LOGITS

<bos>

1.73

Siapa

1.48

 alltså

1.48

 ktorí

1.48

atória

1.48

źć

1.46

peteer

1.42

 mycket

1.41

他们的

1.41

ebly

1.41

Activations Density 0.034%