INDEX

Explanations

semicolon lists

The neuron responds to semicolon punctuation (especially list‐separating “;” tokens).

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

.'

-1.66

orianCalendar

-1.65

="

-1.49

雳

-1.48

郸

-1.47



-1.46

nóstico

-1.46

]))

-1.44

enedig

-1.43

ство

-1.42

POSITIVE LOGITS

dags

1.63

of

1.63

的笑容

1.55

panasonic

1.55

brak

1.53

liu

1.50

 πρ

1.50

ᐛ

1.43

 iſt

1.42

 giardino

1.42

Activations Density 0.035%