INDEX

Explanations

code blocks and specific phrases

closing curly braces or brackets in code that end code blocks or data structures.

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

蚰

-1.92

Iceland

-1.91

 Wszyst

-1.76

 pertenec

-1.74

 aquellos

-1.72

 embarazada

-1.72

mpagne

-1.70

戕

-1.67

 ktorí

-1.67

 býval

-1.66

POSITIVE LOGITS

攆

2.52

媖

2.30

2.23

no

2.19

ၟ

2.11

2.06

ဴ

2.06

which

2.06

</strong>

1.98

ner

1.95

Activations Density 0.000%