INDEX

Explanations

wanted toworks great

This neuron fires on user-expressed desires or enthusiastic endorsements—e.g. “always wanted to…” or “works great.”

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

 одного

-1.42

 动漫

-1.32

҃

-1.29

 ワゴン

-1.29

껏

-1.28

⽼

-1.27

頽

-1.27

行った

-1.26

曖昧さ

-1.22

véd

-1.21

POSITIVE LOGITS

1.37

AttributedString

1.32

 they

1.32

->

1.30

 seems

1.24

but

1.22

atihan

1.20

putar

1.20

全

1.20

 quedando

1.19

Activations Density 0.132%