INDEX

Explanations

measurement units for physical properties

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

-0.12

-0.11

-0.10

yyyy

-0.10

ãĥ¥

-0.10

yyy

-0.09

ÑģÑĮ

-0.09

agrid

-0.09

edList

-0.09

POSITIVE LOGITS

eters

0.17

Ã¨tres

0.17

trs

0.16

illion

0.14

ÙĬÙĦØ§Ø¯

0.14

xmm

0.13

'er

0.13

elsius

0.13

iles

0.12

ilion

0.12

Activations Density 0.072%