INDEX

Explanations

varied domain-specific nouns

This neuron fires on plural nouns that name categories or sections in enumerated lists (e.g. “images,” “videos,” “nursing,” “dentistry,” “veterinary,” “bars,” “cafes,” etc.).

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

 desarrollando

-1.30

 préfère

-1.30

 creciendo

-1.26

 veck

-1.20

:[{

-1.18

 detalj

-1.16

 lamang

-1.13

 wiedzy

-1.11

bedaan

-1.10

他們

-1.09

POSITIVE LOGITS

 this

1.88

but

1.66

 including

1.61

 should

1.55

as

1.37

It

1.34

are

1.31

 termasuk

1.28

 according

1.27

including

1.26

Activations Density 0.544%