INDEX

Explanations

describing qualities or characteristics

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

 brat

-0.10

 prostituer

-0.10

upil

-0.09

ishly

-0.09

 dicks

-0.09

 Dunn

-0.09

ìŀĲìĿ¸

-0.09

upp

-0.09

eting

-0.09

 heroine

-0.08

POSITIVE LOGITS

 little

0.18

 thing

0.14

 ones

0.13

little

0.12

 person

0.12

 piece

0.12

 ingen

0.11

guy

0.11

 excuse

0.10

oref

0.10

Activations Density 0.090%