INDEX

Explanations

The neuron fires on multi-digit numeric tokens (e.g. identifiers or patent/class numbers).

numbers after "msg" or "D"

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

-1.16

З

-1.10

-1.07

«

-1.05

 второго

-1.05

日閲覧

-1.04

۵

-1.04

my

-1.02

 their

-1.02

POSITIVE LOGITS

Wady

1.41

Passer

1.25

 اینترن

1.18

Selamat

1.18

Ingredi

1.18

 mengen

1.13

Publik

1.12

 ognuno

1.12

Ár

1.10

Fonti

1.10

Activations Density 0.042%