INDEX

Explanations

by the day since natural

The neuron fires on words that express a speaker’s attitude or judgment—e.g. opinion, evaluation, or suggestion verbs/adverbs like “seem,” “preferably,” “interesting,” “notice,” and “read up.”

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

</em>

-1.35

ఛ

-1.16

in

-1.10

 bazie

-1.10

 Bias

-1.09

Laser

-1.09

 اذا

-1.05

２

-1.05

يتم

-1.05

 よう

-1.05

POSITIVE LOGITS

trui

1.54

 Therefore

1.44

 therefore

1.37

ܙ

1.33

色々と

1.33

 Hence

1.25

 Selain

1.24

ܐ

1.23

why

1.23

吲

1.23

Activations Density 0.737%