INDEX

Explanations

introduction section headings

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

 colegas

1.35

 নিতান্ত

1.32

ったら

1.31

 gosta

1.29

 cómod

1.26

ड्स

1.25

ählten

1.24

 menos

1.24

 दिनों

1.24

 ziemlich

1.23

POSITIVE LOGITS

 featuring

1.43

 with

1.32

 containing

1.32

 consisting

1.27

azine

1.18

for

1.18

 incorporating

1.13

 allowing

1.10

 comprising

1.10

 enabling

1.06

Activations Density 1.048%