INDEX

Explanations

describing a state, shape, or nuance

The neuron fires on mentions of something’s “state” (e.g. alignment state, exposed state, dry state, loading state) in technical descriptions.

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

跸

-2.22

鹪

-1.98

蒌

-1.98

 帅气

-1.95

 Профиль

-1.91

 Сайт

-1.89

蛲

-1.86

bluza

-1.84

鼍

-1.84

sprzedam

-1.82

POSITIVE LOGITS

ca

1.86

not

1.76

</h4>

1.68

 each

1.67

 albeit

1.66

ed

1.66

me

1.63

 everyone

1.59

↑↑↑</

1.58

al

1.57

Activations Density 0.011%