INDEX

Explanations

references to specific classes or modules within an API or programming environment

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

 handle

-0.74

 handling

-0.71

 handled

-0.69

handle

-0.68

 Handling

-0.67

 handles

-0.66

 Handle

-0.65

Handle

-0.64

handled

-0.62

handling

-0.61

POSITIVE LOGITS

 manager

2.52

 Manager

2.50

Manager

2.47

manager

2.39

 MANAGER

2.30

 Managers

2.20

 managers

2.06

MANAGER

1.98

Managers

1.92

managers

1.80

Activations Density 0.031%