INDEX

Explanations

Homeland Security and Interior Ministry

The neuron detects mentions of government internal‐affairs or security departments (e.g. “Interior,” “Home Affairs,” “Homeland Security”).

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

땃

-1.54

 sprawie

-1.54

-1.50

寕

-1.45

鲠

-1.45

菒

-1.45

</td>

-1.43

 角色

-1.39

 IBRARY

-1.39

蹰

-1.38

POSITIVE LOGITS

 then

1.49

corpions

1.30

at

1.25

⢊

1.25

 first

1.23

Estás

1.22

1.21

 initially

1.20

 haven

1.19

Activations Density 0.018%