INDEX

Explanations

assure, assurance, assured

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

 Warning

-0.10

 warning

-0.09

arter

-0.09

exc

-0.09

apol

-0.09

 Cary

-0.09

Warning

-0.09

 warnings

-0.09

opsis

-0.08

reten

-0.08

POSITIVE LOGITS

 reass

0.41

assic

0.33

ass

0.30

ass

0.29

 reassure

0.28

 assurance

0.28

 Assurance

0.27

Ass

0.25

 comfort

0.25

Ass

0.23

Activations Density 0.083%