INDEX

Explanations

numbers, quantities, counting

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

 Eleven

-0.11

 Ð²ÑģÑİ

-0.08

ãģļ

-0.08

Various

-0.08

 ÑĢÐ°Ð·Ð»Ð¸ÑĩÐ½ÑĭÐµ

-0.08

Dre

-0.08

 Various

-0.08

.:.

-0.08

 meses

-0.07

-plus

-0.07

POSITIVE LOGITS

 three

0.47

two

0.44

 four

0.42

ä¸īä¸ª

0.39

 five

0.39

ä¸¤ä¸ª

0.32

three

0.32

six

0.31

two

0.30

 drei

0.29

Activations Density 0.617%