INDEX

Explanations

comprehensive breakdown

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 papers

0.90

 speeches

0.83

 концер

0.81

 autographs

0.74

 পত্রিকাটি

0.74

 interviews

0.71

 documents

0.69

 newspapers

0.69

 Papers

0.68

 entries

0.68

POSITIVE LOGITS

 detailed

1.33

detailed

1.27

 comprehensive

1.24

 thorough

1.22

 quick

1.16

 brief

1.12

 rundown

1.09

 overview

1.09

 exhaustive

1.07

 breakdown

1.06

Activations Density 0.558%