INDEX

Explanations

punctuation followed by capitalized words

The neuron specializes in spotting bibliographic reference markers—i.e. the periods and numeric codes that delimit and index entries in a reference/citation list.

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

 คา

-1.62

久しぶりの

-1.50

*************/

-1.48

 Hecho

-1.48

 phénom

-1.45

❕

-1.43

hâ

-1.41

㍑

-1.41

 チラ

-1.41

 Fuerza

-1.41

POSITIVE LOGITS

2.05

1.68

not

1.63

1.60

in

1.46

so

1.43

to

1.38

 Does

1.32

 global

1.32

out

1.32

Activations Density 0.002%