INDEX

Explanations

AI self-description

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

jog

-0.10

zan

-0.09

uplic

-0.09

Jog

-0.09

led

-0.09

agon

-0.09

.unbind

-0.09

-0.08

 Claw

-0.08

 conj

-0.08

POSITIVE LOGITS

 exist

0.16

 existence

0.13

ä¸įåŃĺåľ¨

0.12

æĺ¯æĪĳ

0.12

Exist

0.11

 exists

0.11

 Exist

0.11

 opport

0.11

 lack

0.10

existence

0.10

Activations Density 0.114%