INDEX

Explanations

numbers and conjunctions

counting numbers and lists

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

̈

0.47

łaszcza

0.43

̂

0.42

temperature

0.41

においても

0.41

^{\

0.41

indeed

0.40

ইতে

0.40

thousands

0.40

̉

0.40

POSITIVE LOGITS

two

1.32

 två

1.16

 deux

1.16

 zwei

1.15

 двумя

1.14

 dvě

1.12

 δύο

1.10

 двох

1.10

 three

1.09

 два

1.08

Activations Density 0.070%