INDEX

Explanations

words which are positive or indicate things are going well

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PathVariable

-0.74

transQ

-0.62

principalColumn

-0.61

 ویکی‌پدیای

-0.61

🔗

-0.60

 wilful

-0.59

ViewFeatures

-0.59

zeera

-0.59

Kuz

-0.59

invokeLater

-0.57

POSITIVE LOGITS

 good

1.59

good

1.54

Good

1.37

 GOOD

1.34

GOOD

1.33

 excellent

1.30

excellent

1.30

 Good

1.26

 bagus

1.25

 quality

1.20

Activations Density 3.001%