INDEX

Explanations

age expressed as number then year

The neuron flags numeric age expressions (e.g., “27,” “29,” “31-year-old”).

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

the

-1.36

 ilang

-1.17

)['

-1.16

荅

-1.16

 素人

-1.13

 jūs

-1.13

اسخ

-1.13

 regiões

-1.12

Fornecedor

-1.12

଴

-1.11

POSITIVE LOGITS

 then

1.54

’

1.32

his

1.27

 удачи

1.24

最后

1.23

“

1.22

vimento

1.20

 участок

1.15

ally

1.13

1.11

Activations Density 0.059%