INDEX

Explanations

proper nouns and end-of-phrase punctuations

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

ï½ľ

-0.11

mts

-0.10

ï¼ģ”;\n\n

-0.10

WithContext

-0.09

aison

-0.09

åŀĤ

-0.08

cona

-0.08

_Osc

-0.08

POSITIVE LOGITS

 Dahl

0.10

ãĢįçļĦ

0.09

_named

0.08

 Bucc

0.08

 ÄĲÃło

0.08

 Cleveland

0.08

ght

0.08

 acceler

0.08

:,

0.08

BOOLE

0.08

Activations Density 0.100%