INDEX

Explanations

understanding relationships

The neuron flags words that introduce questions or concerns—i.e. discourse markers like “raises,” “questions,” “concerning,” “how,” “about,” etc., used when framing inquiries or issues.

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

šo

-1.21

 this

-1.20

câ

-1.18

 these

-1.17

 musisz

-1.16

 eivät

-1.11

这一幕

-1.10

 namelijk

-1.10

çok

-1.09

 embalaje

-1.08

POSITIVE LOGITS

it

1.53

それを

1.51

 broader

1.38

autres

1.30

 thereof

1.29

—

1.27

 they

1.25

to

1.24

 others

1.23

 associated

1.21

Activations Density 0.160%