INDEX

Explanations

adjective followed by noun

The neuron spotlights strong intensifying adjectives and adverbs that emphasize importance, degree, or similarity (e.g. “vital,” “strategically,” “very,” “important,” “larger,” “lively”).

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

 what

-2.11

 would

-1.57

ternut

-1.50

why

-1.44

もありました

-1.43

 when

-1.41

Při

-1.40

ogenesis

-1.39

蹓

-1.38

setColumns

-1.37

POSITIVE LOGITS

 bernama

1.54

ほとん

1.40

 Fifteen

1.39

 новом

1.38

簠

1.36

 великолеп

1.35

 melainkan

1.35

dosen

1.34

 edifício

1.33

 absence

1.32

Activations Density 0.241%