INDEX

Explanations

producing results

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

ilangkan

-1.01

ignements

-0.92

 speelt

-0.87

žin

-0.85

頜

-0.84

 mengganggu

-0.80

销售

-0.80

 поможет

-0.78

 även

-0.77

 помогает

-0.77

POSITIVE LOGITS

 produce

4.69

 producing

4.31

 produces

4.22

 Produce

3.55

produce

3.55

 yields

3.48

 yielding

3.48

producing

3.42

 produced

3.39

 yield

3.39

Activations Density 0.410%