INDEX

Explanations

numbers followed by parentheses

The neuron primarily detects standalone numeric tokens (numbers) 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

in

-1.45

他说

-1.23

ceptable

-1.22

 voren

-1.20

曬

-1.18

vivo

-1.16

county

-1.16

 cubos

-1.15

鬓

-1.13

POSITIVE LOGITS

タイプの

1.58

烀

1.47

opheles

1.34

 Schlafzimmer

1.31

↵↵↵↵↵↵↵↵↵↵↵↵

1.30

↵↵↵↵↵↵↵↵↵↵↵

1.29

盌

1.27

bộ

1.27

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.24

 würde

1.23

Activations Density 0.004%