INDEX

Explanations

numbers and quantities

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

-0.09

AGAIN

-0.08

á»¯a

-0.08

ãĢĪ

-0.08

 YÃ¶n

-0.08

 ÙĩØ²Ø§Ø±

-0.08

ä¸ĢçĤ¹

-0.08

 Rarity

-0.08

POSITIVE LOGITS

 number

0.41

 series

0.38

 several

0.35

few

0.33

 aantal

0.33

 range

0.32

 variety

0.30

 serie

0.30

ç³»åĪĹ

0.28

 couple

0.26

Activations Density 0.282%