INDEX

Explanations

Numbers written out

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 digitais

-1.05

 varandra

-1.01

%?

-0.96

 }}"></

-0.96

SequentialGroup

-0.96

OGND

-0.94

 démocr

-0.94

 chaude

-0.93

Datuak

-0.93

fjspx

-0.93

POSITIVE LOGITS

 five

0.58

six

0.56

 nine

0.55

 seven

0.55

 four

0.54

ten

0.53

 eight

0.51

two

0.47

one

0.46

 three

0.44

Activations Density 0.001%