INDEX

Explanations

code, comments, copyright

single-character variable names or identifiers in programming 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

Italie

-1.60

そうですが

-1.46

both

-1.44

يًا

-1.41

Largeur

-1.39

modellen

-1.38

étend

-1.37

wanzig

-1.37

ittens

-1.37

 間取り

-1.35

POSITIVE LOGITS

who

1.77

).

1.66

 ebenfalls

1.55

 чтобы

1.46

 Then

1.46

 который

1.44

 что

1.44

}$.

1.40

↵

1.39

Activations Density 0.001%