INDEX

Explanations

explaining concepts, historical accounts

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

他又

1.35

他也

1.26

 told

1.24

γγελ

1.23

 insisted

1.20

 once

1.18

我又

1.18

 gentlemen

1.18

 instructed

1.13

 penned

1.13

POSITIVE LOGITS

 পরিমাণে

1.27

น้ำ

1.23

 Рассмотрим

1.22

FloatTensor

1.18

 Восто

1.17

пры

1.17

ключа

1.16

ить

1.15

ប្រភេទ

1.15

 აღ

1.15

Activations Density 0.015%