INDEX

Explanations

as far as ... is concerned

class definitions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 ancak

0.45

but

0.43

 tellement

0.43

 لیکن

0.43

 pero

0.42

लेकिन

0.42

 اما

0.41

另外

0.41

 fakat

0.40

 mutta

0.40

POSITIVE LOGITS

而言

0.71

来说

0.66

來說

0.61

为例

0.58

来讲

0.55

{};

0.48

来看

0.46

φορά

0.44

相比

0.42

break

0.42

Activations Density 0.008%