INDEX

Explanations

code keywords and syntax

the presence of numeric literals (numbers) in the code.

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

 hade

-1.66

鸲

-1.56

巂

-1.55

began

-1.51

taient

-1.50

囖

-1.49

ﺪ

-1.48

ebly

-1.47

ᒍ

-1.47

nadequate

-1.46

POSITIVE LOGITS

and

2.33

of

2.09

 like

1.78

to

1.77

now

1.52

 with

1.41

no

1.37

 constantly

1.33

They

1.31

out

1.30

Activations Density 0.064%