INDEX

Explanations

surnames and company names

This neuron activates on personal names or usernames (e.g. author signatures) in the text.

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

of

-1.80

чя

-1.67

 таких

-1.53

(!__

-1.41

 такого

-1.40

роме

-1.37

 hende

-1.35

They

-1.31

这种

-1.31

POSITIVE LOGITS

any

1.52

 škoda

1.45

one

1.42

当时的

1.40

 detaylı

1.36

 EVERY

1.29

 what

1.28

ご存知

1.23

 every

1.21

íny

1.21

Activations Density 0.108%