INDEX

Explanations

parenthetical description followed by response

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

 vocês

1.31

 fraîche

1.23

 férias

1.22

kā

1.19

 kiló

1.19

 próximos

1.17

 déterminer

1.14

 excitedly

1.14

 Müller

1.11

 súper

1.11

POSITIVE LOGITS

Also

1.02

 Also

1.02

Ш

0.96

also

0.95

only

0.92

raded

0.90

rayed

0.90

विरुद्ध

0.90

Нет

0.89

reduced

0.87

Activations Density 0.524%