INDEX

Explanations

XML attributes and configuration

The neuron fires on numeric literal tokens (sequences of digits) in structured documents.

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

 their

-2.20

any

-1.42

as

-1.40

an

-1.38

or

-1.31

 they

-1.27

our

-1.22

 голове

-1.18

 Many

-1.16

 with

-1.15

POSITIVE LOGITS

 contextLoads

1.52

は

1.48

criture

1.42

 clouded

1.32

を持ち

1.31

 pido

1.29

がありますが

1.28

zenden

1.27

？』

1.26

itars

1.23

Activations Density 0.003%