INDEX

Explanations

declaring variables and functions

The neuron fires on fragments of programming identifiers or names that mix uppercase letters (and sometimes digits), e.g. parts of CamelCase or numeric identifiers.

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

ждают

-1.16

〼

-1.13

 genieten

-1.13

meisters

-1.09

Zitat

-1.09

 muualla

-1.07

 sostu

-1.06

つづく

-1.06

有些人

-1.06

 maniere

-1.05

POSITIVE LOGITS

 from

1.66

<bos>

1.63

by

1.46

 that

1.39

 cómo

1.25

 only

1.22

()

1.20

();

1.19

אם

1.17

how

1.15

Activations Density 0.022%