INDEX

Explanations

citation identifiers and section titles

The neuron fires on numeric strings (digits and numbers), especially those found in citations, DOIs, section markers, and URLs.

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

of

-1.30

im

-1.29

...

-1.29

OF

-1.21

-1.13

 gren

-1.11

me

-1.10

 THAT

-1.10

 auch

-1.09

POSITIVE LOGITS

معرفی

1.72

Komentar

1.70

توضیحات

1.66

Tekst

1.60

Tentang

1.59

Leírás

1.57

Penulis

1.54

Ukuran

1.50

Zdra

1.48

Catatan

1.47

Activations Density 0.022%