INDEX

Explanations

Gemini models and architecture

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ties

1.37

ā

1.29

 rumored

1.25

ting

1.23

ึ่ง

1.23

taus

1.22

 rumors

1.21

 வெளியான

1.21

ines

1.20

the

1.19

POSITIVE LOGITS

zelfde

1.66

ки

1.56

ى

1.52

д

1.41

ларга

1.40

ми

1.39

ك

1.39

ו

1.39

み

1.39

 tantalum

1.36

Activations Density 0.006%