INDEX

Explanations

text understanding and translation

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

 méthod

1.48

 stealth

1.26

 yoke

1.25

 infiltr

1.25

Transition

1.24

 reversion

1.23

 regroup

1.23

 bifur

1.23

 поверну

1.22

 присоеди

1.19

POSITIVE LOGITS

works

1.55

 literature

1.41

 сказа

1.38

 Scriptures

1.35

렀

1.30

literature

1.30

 LITERATURE

1.30

Works

1.29

 works

1.27

 Scripture

1.25

Activations Density 0.408%