INDEX

Explanations

used with cartridges

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

Sensitive

-0.11

wrong

-0.11

 reused

-0.10

compact

-0.10

-sensitive

-0.10

stu

-0.10

available

-0.10

 sensitive

-0.10

 cheap

-0.10

Bre

-0.10

POSITIVE LOGITS

NON

0.12

 escap

0.11

 Stable

0.11

 strictly

0.10

 uncont

0.10

non

0.10

dep

0.10

 stable

0.10

 benign

0.10

enco

0.09

Activations Density 0.078%