INDEX

Explanations

title is description

This neuron activates on sentences that begin with a title or named work (book, album, film, etc.) functioning as the grammatical subject immediately followed by a verb (e.g. “is,” “offers,” “demonstrates”).

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

それぞれ

-1.51

 These

-1.48

to

-1.39

これらの

-1.37

这些

-1.37

 swoich

-1.27

それぞれの

-1.19

 pourrais

-1.16

 falsos

-1.16

 และ

-1.16

POSITIVE LOGITS

 successivo

1.44

but

1.39

any

1.32

 various

1.24

 specified

1.23

 câmera

1.21

 чёрный

1.20

 mentioned

1.19

 presentato

1.18

But

1.17

Activations Density 0.270%