INDEX

Explanations

house arrest, restrictions, unable

The neuron strongly activates on words referring to being confined or detained (e.g. arrest, jail, exile).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Lazy

-0.94

CEED

-0.91

lapia

-0.90

wsze

-0.90

abandon

-0.90

Destroyed

-0.87

λοι

-0.85

 Escolar

-0.85

 Mahl

-0.84

ޫ

-0.83

POSITIVE LOGITS

 restricted

1.60

 unable

1.55

 restrictions

1.49

 restriction

1.45

 cannot

1.36

restricted

1.34

 constraint

1.34

只能

1.34

限制

1.29

 powerless

1.27

Activations Density 0.057%