INDEX

Explanations

phrases that describe something being impossible or difficult to achieve

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

GOTREF

-0.79

oneofs

-0.64

+#+

-0.62

 uska

-0.60

 iLocal

-0.58

Datuak

-0.53

mpg

-0.52

 numberWith

-0.52

timeScale

-0.52

#+#

-0.51

POSITIVE LOGITS

 bounds

0.62

 logistique

0.61

 Beyond

0.59

 BEYOND

0.59

Within

0.58

Beyond

0.58

/*

0.57

 confines

0.57

 beyond

0.56

 Within

0.56

Activations Density 0.108%