INDEX

Explanations

common words like "the", "a", "of", "and" alongside words related to people, especially "user".

the

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

awtextra

-0.75

 ModelExpression

-0.68

 jScrollPane

-0.65

+#+#

-0.64

enterOuterAlt

-0.64

delwed

-0.63

StructEnd

-0.62

 astore

-0.61

MessageOf

-0.60

LEncoder

-0.59

POSITIVE LOGITS

Datuak

0.72

ISupport

0.52

 purpoſe

0.50

Tad

0.50

 ſen

0.48

 teams

0.48

 pleaſure

0.48

Tad

0.47

 minority

0.46

 luận

0.45

Activations Density 1.462%