INDEX

Explanations

package, Cyclic, Tuesday

This neuron primarily detects sentence-ending punctuation (especially periods marking the end of a sentence).

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

and

-2.00

-1.51

 Jednak

-1.28

 деца

-1.27

 tambien

-1.26

millimeters

-1.20

 الاجتماعية

-1.20

of

-1.20

 حوالي

-1.19

</strong>

-1.16

POSITIVE LOGITS

the

2.08

 other

1.61

 penghar

1.57

 habrían

1.56

 these

1.55

 advirtió

1.54

 maksud

1.52

 manfaat

1.51

 digitais

1.46

uszka

1.45

Activations Density 0.028%