INDEX

Explanations

evaluative adjectives + common nouns

The neuron lights up on evaluative words and phrases carrying strong sentiment or judgment (e.g. “good job,” “sure,” “great,” “tough place,” “flown by”).

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

who

-1.43

ଃ

-1.42

 evtl

-1.33

 ferie

-1.30

probably

-1.29

阄

-1.29

ೢ

-1.27

 gewi

-1.27

乀

-1.27

 magari

-1.24

POSITIVE LOGITS

 changes

1.47

 other

1.43

変わる

1.35

 enjo

1.33

providedIn

1.29

 equipe

1.27

 perubahan

1.22

 intricate

1.20

 there

1.16

夏休み

1.16

Activations Density 0.057%