INDEX

Explanations

become, became, becoming

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

Negative Logits

 Famous

-0.10

IFO

-0.09

 famous

-0.09

INGER

-0.09

undef

-0.09

 prestigious

-0.08

ï¼«

-0.08

/on

-0.08

_PIX

-0.08

POSITIVE LOGITS

 become

0.29

 became

0.25

 becoming

0.25

æĪĲä¸º

0.25

 menjadi

0.24

 Become

0.23

 becomes

0.22

 trá»Ł

0.22

bec

0.21

æĪĲçĤº

0.21

Activations Density 0.152%

No Known Activations