INDEX

Explanations

phone numbers

The neuron fires on numeric tokens (digits or numbers) such as phone‐number or address digits.

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

 hvil

-1.41

 because

-1.35

 ocultar

-1.31

 münchen

-1.29

 fides

-1.27

 alpes

-1.27

 bús

-1.26

 Kette

-1.25

 admitir

-1.23

-------

-1.23

POSITIVE LOGITS

不僅

1.28

 Фі

1.24

Seen

1.24

 населення

1.23

TO

1.23

 dicha

1.20

attaque

1.20

escalier

1.16

Via

1.16

MA

1.16

Activations Density 0.027%