INDEX

Explanations

content from research papers, including in-text citations, the word "study", and section headings

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IsContent

-0.88

 staffel

-0.77

 pleaſure

-0.77

 auffi

-0.73

^(@)

-0.71

 propOrder

-0.70

Efq

-0.68

 myſelf

-0.68

{?>

-0.66

Портал

-0.66

POSITIVE LOGITS

diğini

0.53

nor

0.52

nor

0.50

roux

0.49

 farlo

0.48

 trụ

0.45

 TextAppearance

0.44

 récol

0.43

 váll

0.41

bő

0.41

Activations Density 0.223%