INDEX

Explanations

the reporting of experimental results in scientific literature

research study results

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

Negative Logits

 فريبيس

-0.55

centralwidget

-0.51

tikzpicture

-0.51

itinéraire

-0.50

 cref

-0.49

 Paglinawan

-0.49

цезда

-0.49

CloseOperation

-0.48

bewerken

-0.48

 snippetHide

-0.48

POSITIVE LOGITS

OGND

0.69

 veneta

0.55

DebuggerNonUser

0.54

WebVitals

0.53

EDEFAULT

0.52

 offending

0.50

ơn

0.49

 refiere

0.48

はじめに

0.48

Referencie

0.47

Activations Density 15.264%