INDEX

Explanations

directing reader's attention

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Efq

-0.69

 mukana

-0.66

Abitanti

-0.66

ADT

-0.61

 pseudonym

-0.60

AndroidJUnit

-0.58

 logar

-0.55

WireFormat

-0.54

 concentric

-0.54

sterdam

-0.54

POSITIVE LOGITS

out

0.98

Out

0.74

it

0.67

ou

0.66

OUT

0.65

 them

0.57

Out

0.54

 متعلقه

0.54

reportWebVitals

0.52

 your

0.49

Activations Density 0.005%