INDEX

Explanations

The neuron activates on numeric tokens (including integers, years, and decimal numbers).

Prime Minister, United States, Islam

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

 الجلد

-1.49

 بيان

-1.48

 گیاه

-1.40

ׇ

-1.39

 овощи

-1.34

ܐ

-1.31

you

-1.28

 الشرطة

-1.27

 النبات

-1.26

About

-1.23

POSITIVE LOGITS

ִּ

1.81

ּוֹ

1.73

וּ

1.45

ְר

1.45

 бампер

1.40

ַּ

1.36

᾽

1.35

ְּ

1.24

 analysts

1.23

 Janeiro

1.20

Activations Density 0.017%