INDEX

Explanations

locations and organizations

The neuron fires on major category‐label nouns—words like “show,” “series,” “service,” “site,” “island,” and “nation” that head descriptive noun phrases.

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

 these

-1.49

這些

-1.46

 campionato

-1.41

Эти

-1.34

 lutego

-1.34

 those

-1.33

 listopada

-1.33

 what

-1.33

 such

-1.30

 them

-1.26

POSITIVE LOGITS

could

1.22

 aseguró

1.21

are

1.20

 podría

1.19

 ihren

1.19

if

1.18

1.17

すでに

1.17

楽しめる

1.15

Activations Density 0.123%