INDEX

Explanations

political leaders and administrations

The main thing this neuron does is detect mentions of political figures, government offices, and related proper names (e.g., presidents, governors, ministries, legislative bodies).

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

at

-1.41

 with

-1.35

if

-1.34

as

-1.34

out

-1.24

芋

-1.20

 because

-1.19

ğın

-1.17

Πηγή

-1.17

 vooral

-1.17

POSITIVE LOGITS

 totalement

1.42

 temporaire

1.37

它们

1.31

 muur

1.29

他们的

1.28

ORETICAL

1.27

 définitivement

1.23

镲

1.20

 местах

1.20

 pantaloni

1.19

Activations Density 0.091%