INDEX

Explanations

how things are usually done

The neuron detects adverbs or adjectives that signal something is customary or typical (e.g. words meaning “usually,” “typically,” or “ordinarily”).

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

และ

-2.13

 மற்றும்

-1.83

and

-1.82

ScopeManager

-1.77

isième

-1.71

ilarities

-1.63

⸒

-1.59

꜆

-1.58

 maravilla

-1.56

rencies

-1.55

POSITIVE LOGITS

re

1.95

of

1.86

1.80

 regarded

1.79

 only

1.79

1.74

Do

1.73

 zuerst

1.70

If

1.70

No

1.70

Activations Density 0.059%