INDEX

Explanations

roles and their names

The neuron detects common nouns that name people’s occupations or social roles (e.g., “instructor,” “housekeeper,” “doctor,” “bell-boy,” “servants,” “singer”).

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

 this

-1.60

 zowel

-1.39

 those

-1.34

 dagelijks

-1.31

olateral

-1.29

 Tuesday

-1.25

 both

-1.23

 ataupun

-1.23

様々

-1.23

ppuden

-1.21

POSITIVE LOGITS

 apenas

1.36

 appena

1.33

 spielen

1.24

are

1.19

 Songtext

1.17

 掃除

1.16

gö

1.15

TOO

1.13

 juste

1.13

 sacerdotes

1.09

Activations Density 0.071%