INDEX

Explanations

technical content

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 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

 svolta

-0.65

 مفص

-0.59

 Савезне

-0.57

tagHelperRunner

-0.51

cyklopedia

-0.50

iecie

-0.50

 للاسماء

-0.50

 kasarigan

-0.49

ffiche

-0.48

 kereszt

-0.47

POSITIVE LOGITS

 Woodford

0.59

abases

0.54

enya

0.52

 jura

0.51

saraba

0.50

amodel

0.49

flashdata

0.48

 OMITBAD

0.47

Que

0.47

Sor

0.47

Activations Density 0.005%