INDEX

Explanations

text within quotation marks, specifically single quotes that might suggest sarcasm or a certain viewpoint along with double quotes

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

 <<<<<<<<<<<<<<

-0.71

InvalidProtocol

-0.70

 ***!

-0.69

 imdb

-0.67

 avoient

-0.66

 Europeo

-0.66

 Popis

-0.65

 canadien

-0.63

 Batal

-0.61

 AssemblyTitle

-0.61

POSITIVE LOGITS

ViewFeatures

0.59

lolo

0.58

nadequate

0.55

 propOrder

0.52

jones

0.52

kenstock

0.51

dega

0.51

สึก

0.49

》。

0.49

émoc

0.48

Activations Density 0.026%