INDEX

Explanations

code syntax, programming keywords

line breaks or newlines in code, especially after comments or at the start of new lines.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

of

-1.73

ことなく

-1.64

ことができ

-1.53

endimiento

-1.45

 discus

-1.39

)}+

-1.35

水面

-1.34

 larg

-1.30

we

-1.30

还有一

-1.29

POSITIVE LOGITS

or

1.97

 each

1.60

 every

1.52

 their

1.51

not

1.51

 themselves

1.45

 this

1.41

 EVERY

1.40

或

1.32

où

1.31

Activations Density 0.088%