INDEX

Explanations

healthy or unhealthy food

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

cule

-0.91

Coffee

-0.79

izacji

-0.74

vul

-0.73

 intensiv

-0.72

Vul

-0.72

mechanical

-0.72

نان

-0.72

coffee

-0.71

marginBottom

-0.71

POSITIVE LOGITS

 healthy

4.75

 healthier

4.28

healthy

3.84

 Healthy

3.63

 health

3.45

Healthy

3.31

 healthful

3.00

 unhealthy

2.92

 healthiest

2.91

 HEALTH

2.89

Activations Density 0.059%