INDEX

Explanations

questions or conditional statements

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

 Roskov

-1.17

ValueStyle

-0.89

AddTagHelper

-0.89

oredCriteria

-0.89

ThroughAttribute

-0.89

">—

-0.88

Personendaten

-0.88

 незавершена

-0.88

 виправивши

-0.85

NameInMap

-0.84

POSITIVE LOGITS

we

0.71

 they

0.56

you

0.50

ain

0.49

 signing

0.46

 people

0.46

 answering

0.45

sprechen

0.45

 Folge

0.45

 this

0.44

Activations Density 0.203%