INDEX

Explanations

conversational starters and continuations

This neuron fires on sentence- or turn-initial discourse markers—especially standalone first-person “I” and similar opening words (e.g. “Well,” “Thanks,” “If”).

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

to

-2.22

by

-2.19

着自己的

-1.63

 with

-1.59

着一

-1.56

了一种

-1.55

着一个

-1.53

出了一

-1.45

of

-1.43

as

-1.43

POSITIVE LOGITS

1.85

 nikt

1.54

 jesteśmy

1.50

 nasz

1.49

Where

1.48

What

1.46

 mantend

1.45

lelő

1.45

 mogę

1.42

 melalui

1.41

Activations Density 0.254%