INDEX

Explanations

code identifiers and paths

code-related terms and identifiers, particularly variable names, file names, HTML/XML tags, and function names in programming contexts.

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

one

-1.80

aruhi

-1.53

 during

-1.39

jedno

-1.32

 months

-1.31

água

-1.27

 what

-1.27

camara

-1.26

 like

-1.26

 sebelum

-1.23

POSITIVE LOGITS

na

1.59

atique

1.53

ur

1.45

 then

1.44

illeurs

1.40

pa

1.39

ícias

1.36

atized

1.32

現在は

1.31

spro

1.30

Activations Density 0.655%