INDEX

Explanations

vary depending on context

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

 varying

-0.13

 uniquely

-0.10

 varied

-0.10

 varies

-0.10

 differently

-0.09

 vary

-0.09

 Vari

-0.09

vary

-0.09

 different

-0.08

 varieties

-0.08

POSITIVE LOGITS

 Hort

0.12

 cheer

0.11

 Goodman

0.10

Zwe

0.10

 legitimately

0.10

/File

0.10

ANTI

0.10

oras

0.10

=id

0.10

 desperation

0.10

Activations Density 0.634%