INDEX

Explanations

scientific/technical language

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 commerciales

-0.79

 migration

-0.78

 colorés

-0.75

IUrlHelper

-0.75

ⓧ

-0.74

 particulières

-0.71

 préfé

-0.71

 avoient

-0.70

 étoient

-0.70

 ouvriers

-0.70

POSITIVE LOGITS

ed

0.58

 fine

0.57

页面存档备份

0.56

 charges

0.53

 فريبيس

0.52

 Fine

0.51

un

0.50

ia

0.49

ver

0.49

 intr

0.49

Activations Density 1.024%