INDEX

Explanations

higher income individuals

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

 Depths

-0.10

 Minimum

-0.09

 Ober

-0.09

oret

-0.08

 millions

-0.08

 impoverished

-0.08

stry

-0.08

Minimum

-0.08

 mÃ¤nner

-0.08

abis

-0.08

POSITIVE LOGITS

 higher

0.22

 older

0.19

 newer

0.18

higher

0.17

 lower

0.16

 larger

0.15

 high

0.15

 certain

0.15

 younger

0.14

older

0.14

Activations Density 0.133%