INDEX

Explanations

Washington, DC

The neuron never fires on any token—it does not detect any pattern.

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

 himſelf

-2.53

 unange

-2.39

 δεν

-2.25

它是

-2.23

 hadapan

-2.20

 resembling

-2.20

:/

-2.17

䊐

-2.17



-2.16

伈

-2.16

POSITIVE LOGITS

觔

2.36

has

2.17

2.06

屛

1.98

.,

1.98

zelfde

1.92

𖡼

1.92

 faisons

1.92

宝物

1.92

 coraz

1.91

Activations Density 0.005%