INDEX

Explanations

organizing things and information

The neuron activates on different forms of the root “organize,” such as organizer, organizing, and organized.

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

we

-3.41

’

-3.23

釤

-2.67

 $^{\

-2.44

眎

-2.41

 cette

-2.34

$^{

-2.31

 höhere

-2.30

 această

-2.25

POSITIVE LOGITS

In

2.48

hörigen

2.45

Виды

2.42

韞

2.39

You

2.38

Миха

2.23

Like

2.22

袮

2.22

Примеры

2.22

_)

2.20

Activations Density 0.009%