INDEX

Explanations

seeking clarification questions

The neuron fires on discourse markers that signal a request for clarification or definition (e.g. “anyway,” “exactly,” “eigentlich”).

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>

-1.56

and

-1.54

</h4>

-1.41

</h5>

-1.30

merksamkeit

-1.23

 SUCH

-1.21

naar

-1.20

 نکته

-1.15

such

-1.13

 CONSIDER

-1.13

POSITIVE LOGITS

 those

1.84

why

1.50

),

1.47

vités

1.43

рия

1.43

 denen

1.43

 their

1.39

 ubiquitous

1.39

,“

1.38

Нет

1.36

Activations Density 0.012%