INDEX

Explanations

contents of containers

The neuron detects words that express containment or enclosure relationships (e.g. “enclose,” “containing,” “filled with,” “of” in “container of”).

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

 baskets

-1.41

 boxes

-1.35

 bags

-1.31

 buckets

-1.27

 bins

-1.26

 bucket

-1.20

 bowl

-1.20

 containers

-1.16

 pots

-1.16

 barrels

-1.16

POSITIVE LOGITS

 contents

1.80

contents

1.46

 inhoud

1.29

 Contents

1.27

Contents

1.19

 containing

1.13

 goodies

1.09

 содержи

1.04

 innehå

1.02

 contained

0.97

Activations Density 0.074%