INDEX

Explanations

The main thing this neuron does is detect numeric literals (numbers, including integers, decimals, and percentages).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

𖥧

-1.61

 zwölf

-1.42

$*

-1.39

 Daß

-1.34

 zijne

-1.30

 progrès

-1.29

悧

-1.29

schermata

-1.27

 muualla

-1.25

 teeming

-1.25

POSITIVE LOGITS

𞤢

1.42

 educativas

1.40

and

1.40

 sehingga

1.39

in

1.38

so

1.36

،

1.35

}}\

1.34

nitř

1.31

ͤ

1.31

Activations Density 0.001%