INDEX

Explanations

are/were followed by quantity

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bruch

1.39

 Platz

1.38

affee

1.36

 पड़ता

1.35

rón

1.33

 ክፍል

1.33

 sucedido

1.33

okon

1.32

ného

1.31

 exceptionnelle

1.31

POSITIVE LOGITS

<bos>

1.86

 similarities

1.70

 люди

1.69

few

1.67

いくつ

1.62

 criminals

1.62

兩個

1.57

 skills

1.57

 عوامل

1.57

 tools

1.55

Activations Density 0.450%