INDEX

Explanations

geopolitical entities and organizations

The neuron is picking out number tokens (numerical figures, especially years or other multi‐digit numbers).

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

 brinde

-1.27

 seguente

-1.06

 soldati

-0.99

 deciso

-0.90

 станов

-0.90

閲覧

-0.90

จริงๆ

-0.88

 görül

-0.88

собен

-0.88

 seguenti

-0.88

POSITIVE LOGITS

 country

1.97

 region

1.27

 United

1.25

country

1.17

 Organisation

1.09

 nation

1.06

quelles

1.05

 Netherlands

1.05

 страны

1.03

 Organization

0.99

Activations Density 0.116%