INDEX

Explanations

aware, conscious, or mindful

The neuron fires on words and short phrases expressing awareness or careful attention (e.g. “aware of,” “mindful,” “conscientious of”).

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

-1.56

ленная

-1.54

How

-1.46

unier

-1.45

 Currently

-1.43

 Either

-1.41

 Properly

-1.38

 Implementing

-1.35

-1.34

POSITIVE LOGITS

 iſt

1.67

 também

1.62

⟝

1.60

<0xD7>

1.60

！

1.58

౩

1.57



1.56

بیش

1.55

 ваши

1.54

 ciebie

1.51

Activations Density 0.012%