INDEX

Explanations

values, lifestyle, religion, compatibility

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Confidence

-0.11

osas

-0.11

 confidence

-0.11

 Trafford

-0.10

 telegram

-0.10

 Exhaust

-0.10

 Capture

-0.10

 emotion

-0.10

rase

-0.10

 flattering

-0.10

POSITIVE LOGITS

 values

0.19

 lifestyle

0.18

 Values

0.17

 religion

0.16

 lifestyles

0.16

 compatibility

0.16

Values

0.16

 compat

0.15

 incompatible

0.15

 compatible

0.15

Activations Density 0.023%

No Known Activations