INDEX

Explanations

names of people

The neuron is specialized in detecting capitalized proper names—especially surnames.

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

as

-1.93

 every

-1.62

 their

-1.62

on

-1.45

 there

-1.41

 some

-1.35

new

-1.34

 each

-1.33

忏

-1.29

 putting

-1.23

POSITIVE LOGITS

亹

1.52

llavero

1.39

isine

1.35

inaison

1.34

 égard

1.33

MANY

1.33

뻣

1.30

啕

1.30

 Millionen

1.29

 énorm

1.29

Activations Density 0.329%