INDEX

Explanations

descriptions following "a"

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

bisk

-1.01

 Mostly

-0.99

 admittedly

-0.97

şak

-0.97

 importation

-0.93

Mostly

-0.91

vielen

-0.91

鷗

-0.89

Spesifikasi

-0.88

Manfaat

-0.85

POSITIVE LOGITS

 imagery

0.99

 scrutiny

0.99

 legislation

0.98

 ADVICE

0.91

 galore

0.90

 signage

0.90

 memorabilia

0.88

様な

0.88

 collateral

0.88

 footage

0.87

Activations Density 0.078%