INDEX

Explanations

what is the result

This neuron detects numeric expressions and values (numbers, decimals, and percentages) in the text.

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

叆

-1.52

</h2>

-1.48

 inapropiados

-1.42

 suicidio

-1.40

cursos

-1.39

 Likewise

-1.37

 within

-1.36

 einzigen

-1.34

cuid

-1.34

 between

-1.31

POSITIVE LOGITS

 mereka

1.83

 они

1.63

why

1.58

 Mereka

1.58

 they

1.58

 их

1.55

romas

1.49

.'

1.43

姽

1.42

 quando

1.41

Activations Density 0.051%