INDEX

Explanations

sender names and email addresses

The neuron fires on the “<” token that introduces email addresses or header fields (e.g. the “<” before an address in forwarded or reply headers).

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

 posando

-1.67

you

-1.60

 hiasan

-1.52

 schuldig

-1.52

 refrigerador

-1.49

 kompatibel

-1.47

-1.46

 poésie

-1.43

自己的

-1.36

لباس

-1.36

POSITIVE LOGITS

micamente

1.45

of

1.42

 unless

1.38

 gempa

1.37

 between

1.37

authier

1.36

OGLYPH

1.35

;.

1.33

 なし

1.33

↵

1.32

Activations Density 0.001%