INDEX

Explanations

gay relationships and naming

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sexist

-0.17

 heterosexual

-0.17

her

-0.15

 sexism

-0.15

 misogyn

-0.15

 heroine

-0.14

 feminists

-0.14

 girl

-0.14

 vaginal

-0.13

 Girl

-0.13

POSITIVE LOGITS

gay

0.26

gay

0.22

Gay

0.21

 drag

0.20

HIV

0.20

Gay

0.19

-gay

0.19

 AIDS

0.19

 Drag

0.18

 cruising

0.17

Activations Density 0.124%