INDEX

Explanations

fair price

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

 legality

-0.09

otal

-0.09

 impunity

-0.08

ESIS

-0.08

dsn

-0.08

peq

-0.08

arshal

-0.08

 immedi

-0.08

 option

-0.08

POSITIVE LOGITS

 fair

0.54

fair

0.45

 reasonable

0.42

 Fair

0.39

Fair

0.38

reasonable

0.34

 fairly

0.32

 faire

0.32

 Reason

0.30

 fairness

0.30

Activations Density 0.094%