INDEX

Explanations

names of places, organizations, and brands

The neuron predominantly fires on very common function words (articles and short prepositions like “the,” “of,” “to,” etc.).

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

 grasas

-1.59

 lámina

-1.38

λον

-1.34

گه

-1.34

 cáps

-1.33

全都

-1.32

 mascarilla

-1.29

 maleta

-1.26

 carreras

-1.25

 вашу

-1.25

POSITIVE LOGITS

our

1.59

in

1.46

 również

1.28

つづく

1.27

更に

1.27

 utilizó

1.25

 recevrez

1.23

Сю

1.22

サウナ

1.22

大量的

1.21

Activations Density 0.083%