INDEX

Explanations

task and completion status

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Bere

-0.86

Selections

-0.85

Paket

-0.82

MenuList

-0.81

Ered

-0.79

 cartridges

-0.78

 tanker

-0.77

ೃ

-0.77

 pillows

-0.77

ticias

-0.75

POSITIVE LOGITS

 task

2.36

 tasks

2.30

 completed

2.14

task

1.84

 completion

1.76

 Task

1.75

Task

1.66

tasks

1.57

 Tasks

1.55

Tasks

1.50

Activations Density 0.033%