INDEX

Explanations

everyone says something

The neuron detects generic‐universal statements—verbs that describe what “everyone” (or “everybody”) does.

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

With

-1.56

Both

-1.47

of

-1.46

'],$

-1.38

Without

-1.37

Through

-1.35

Those

-1.31

These

-1.30

 where

-1.30

During

-1.27

POSITIVE LOGITS

it

1.48

 their

1.44

⁸

1.27

⎆

1.25

 formules

1.23

 lakini

1.20

tition

1.20

 verhind

1.16

iets

1.15

 knex

1.14

Activations Density 0.102%