INDEX

Explanations

abstract concepts followed by punctuation

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

etc

1.47

and

1.31

and

1.30

but

1.21

1.19

 plus

1.18

<eos>

1.17

и

1.17

etc

1.14

và

1.13

POSITIVE LOGITS

ပါတယ်။

1.52

abbanti

1.49

 살펴보도록

1.48

ಬ್ಬಿಣ

1.48

BottomOf

1.46

 commemoration

1.45

 صہیون

1.45

 다음과

1.44

MSFont

1.43

 namely

1.42

Activations Density 0.542%