INDEX

Explanations

I

The neuron fires on sentence-openers that use discourse markers or first-person pronouns (e.g. “I,” “So,” “How”) at the start of a new line.

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

-1.56

aptors

-1.50

not

-1.48

in

-1.41

metres

-1.36

 leão

-1.33

ｎ

-1.32

ｕ

-1.30

SELF

-1.30

њ

-1.30

POSITIVE LOGITS

 procedentes

1.41

 クリック

1.37

 Then

1.35

 destinées

1.34

1.30

鷂

1.30

Übersetzung

1.29

Witam

1.28

dịch

1.27

INTRODUCCIÓN

1.25

Activations Density 0.037%