INDEX

Explanations

make up for

The neuron activates on the multiword expression “make up” (including variants like “made up”), that is, instances of that verb–particle construction.

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

 seleccionadas

-1.30

鹛

-1.27

 темы

-1.26

௦

-1.24

 gebeten

-1.22

 manuales

-1.20

￡

-1.19

網

-1.19

 purport

-1.18

芸

-1.17

POSITIVE LOGITS

“

1.70

1.28

 When

1.24

℣

1.24

ir

1.22

up

1.17

“

1.13

افية

1.12

 обеспе

1.10

 During

1.09

Activations Density 0.013%