INDEX

Explanations

HTML heading tags

This neuron activates on HTML heading tags (e.g., the “h3” or “h4” tokens), identifying section headers in the markup.

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

-1.27

an

-1.25

 enormous

-1.19

 Beiträge

-1.16

and

-1.15

 having

-1.15

كات

-1.13

 varied

-1.13

 Gonna

-1.12

 essentially

-1.09

POSITIVE LOGITS

러나

1.40

¹,

1.38

ellees

1.38

ας

1.30

 ۱۳

1.30

⁷

1.29

 ansik

1.27

☫

1.27

 могут

1.27

พ

1.27

Activations Density 0.001%