INDEX

Explanations

World of Warcraft game terms

The neuron is primarily firing on numeric tokens (standalone digits, time stamps, counts, IDs) and their adjoining punctuation.

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

 creato

-1.42

 precin

-1.41

ὕ

-1.37

出场

-1.29

ἵ

-1.28

 reda

-1.27

 protege

-1.24

ΠΑ

-1.23

 Gohan

-1.20

 masculina

-1.20

POSITIVE LOGITS

 mantuvo

1.34

 estableció

1.29

for

1.22

 ложки

1.12

当时

1.09

хта

1.08

 utilizó

1.07

at

1.05

明显

1.05

to

1.03

Activations Density 0.038%