INDEX

Explanations

looking for inspiration

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

unp

-0.11

èŁ

-0.09

 tact

-0.09

illac

-0.09

 causing

-0.09

ï½Ľ

-0.08

 propelled

-0.08

ãĥĭãĤ¢

-0.08

spe

-0.08

 sensitive

-0.08

POSITIVE LOGITS

 inspiration

0.65

 inspir

0.56

 Inspiration

0.52

Insp

0.40

 inspire

0.37

 inspired

0.33

 Inspir

0.33

 Ð²Ð´

0.32

 inspires

0.30

 inspirational

0.29

Activations Density 0.157%