INDEX

Explanations

The neuron detects organization names and URLs in software license/copyright headers.

domain names and emails

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

 last

-1.43

 because

-1.41

and

-1.40

 gets

-1.34

之前

-1.29

 męski

-1.26

tidig

-1.25

 almohada

-1.25

 mascarilla

-1.24

in

-1.22

POSITIVE LOGITS

 plupart

1.39

ρ

1.33

 Dinas

1.32

Елена

1.32

\'

1.30

 ...,

1.28

 distanciation

1.26

 SwitchCompat

1.25

',

1.25

 作り方

1.24

Activations Density 0.262%