INDEX

Explanations

street addresses

The neuron activates on numeric tokens—especially the individual digits that make up multi-digit numbers such as years or street addresses.

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

archiviato

-1.44

yship

-1.41

webElement

-1.37

hoku

-1.34

KEYCODE

-1.32

 since

-1.32

⾞

-1.30

since

-1.28

celotti

-1.28

 iniziale

-1.27

POSITIVE LOGITS

 estuviera

1.32

of

1.31

 gyven

1.29

 بنابراین

1.27

 العامة

1.20

be

1.19

.../

1.17

 dezelfde

1.16

STAND

1.16

enten

1.16

Activations Density 0.043%