INDEX

Explanations

инструмент, модели, способов

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

 Фургал

3.46

请

3.32

 охра

3.31

hetiti

3.27

 Цуки

3.23

 осозна

3.23

 оборудование

3.19

请

3.18

 أغ

3.15

Antaeotricha

3.15

POSITIVE LOGITS

1.62

1.28

1.22

of

1.22

1.20

1.19

in

1.10

for

1.09

–

1.08

1.01

Activations Density 0.018%