INDEX

Explanations

the word "control" in the context of either permission or experimentation

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

 hike

-0.59

Sis

-0.59

monk

-0.59

PageContext

-0.58

 Coffey

-0.58

-0.57

 Jennings

-0.57

 Mendez

-0.57

Fonti

-0.56

 rump

-0.56

POSITIVE LOGITS

 control

2.09

 CONTROL

2.05

 Control

1.98

control

1.89

Controlling

1.86

 controls

1.85

Control

1.84

CONTROL

1.80

 Controlling

1.80

 Controls

1.74

Activations Density 0.312%