INDEX

Explanations

most recorded, read, published, visited

The neuron detects quantitative or statistical superlatives and intensity markers (e.g. “most,” “widely,” “average,” “well …”).

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

the

-1.42

-1.36

my

-1.36

-1.30

but

-1.30

re

-1.30

-1.28

my

-1.27

-1.26

POSITIVE LOGITS

迨

1.55

 夜景

1.30

 クロス

1.28

 дорогой

1.27

 normativa

1.24

música

1.23

 materie

1.21

файл

1.21

 AGRIC

1.20

 イベント

1.17

Activations Density 0.123%