INDEX

Explanations

sensitive / sensitive to

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enk

-0.11

licant

-0.10

 hearty

-0.10

kuk

-0.10

 fixing

-0.10

 Muss

-0.09

 plentiful

-0.09

ippi

-0.09

 fixes

-0.09

.eng

-0.09

POSITIVE LOGITS

 sensitive

0.23

 sensitivity

0.23

 delicate

0.23

Sensitive

0.21

 sensit

0.21

 delic

0.20

 fragile

0.18

-sensitive

0.18

ensitive

0.17

 vulnerable

0.16

Activations Density 0.152%