INDEX

Explanations

in the way a

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

 определён

2.44

めっちゃ

2.26

вёр

2.23

 দেওয়ার

2.18

 тяжё

2.17

щё

2.12

 Всё

2.10

めちゃ

2.10

uș

2.06

 সবসময়

2.05

POSITIVE LOGITS

 recognised

1.45

and

1.43

of

1.37

 recognises

1.28

1.27

 realised

1.27

şi

1.24

 tunnelling

1.22

 travellers

1.20

 whilst

1.19

Activations Density 0.066%