INDEX

Explanations

married couples or relationships

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 sequ

-0.08

 bedside

-0.08

olis

-0.08

 Sabbath

-0.07

 Harlem

-0.07

sequ

-0.07

 Cass

-0.07

 elabor

-0.07

rel

-0.07

/print

-0.07

POSITIVE LOGITS

 couples

0.14

 couple

0.13

/single

0.12

 Couple

0.12

uche

0.11

 divorced

0.11

 Bliss

0.11

asin

0.11

 Minute

0.11

-looking

0.10

Activations Density 0.008%

No Known Activations