INDEX

Explanations

references to attributes labeled as "title" in a structured format

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 MonoBehaviour

-0.51

 jsPsych

-0.49

Lax

-0.49

üyada

-0.48

 Jacob

-0.47

monk

-0.47

 Greg

-0.45

 monks

-0.45

 Vans

-0.44

SaveChangesAsync

-0.44

POSITIVE LOGITS

 title

1.32

 getTitle

1.31

title

1.27

Title

1.26

 Title

1.24

getTitle

1.23

 TITLE

1.19

 Titel

1.06

TITLE

1.03

titles

1.02

Activations Density 0.034%