INDEX

Explanations

technical terms with punctuation

This neuron detects structural or “markup” tokens in technical or academic text—e.g. XML tags, section identifiers, citation markers, code selectors, and other formatting elements.

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

was

-1.69

 only

-1.68

to

-1.67

 when

-1.55

 />";

-1.53

it

-1.49

 berle

-1.47

 then

-1.45

 embri

-1.38

 caminha

-1.35

POSITIVE LOGITS

 zarówno

1.37

1.36

 Олег

1.31

 thoſe

1.31

 вчера

1.29

 misschien

1.24

 nowego

1.24

 dzisiaj

1.23

 llenos

1.23

Activations Density 0.123%