INDEX

Explanations

coexistence and living arrangements

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

 collectively

-0.10

 intimate

-0.09

_ISO

-0.09

 coincidence

-0.09

 dood

-0.09

.website

-0.08

ë§ģ

-0.08

arma

-0.08

hun

-0.08

 Naming

-0.08

POSITIVE LOGITS

co

0.40

 living

0.30

 lived

0.25

living

0.24

 live

0.24

 harmon

0.24

 Living

0.23

 harmony

0.22

 conv

0.22

Living

0.21

Activations Density 0.095%