INDEX

Explanations

prepositions and foreign words

The neuron fires on programming‐ or code‐related terms (e.g. “based,” “in,” “executing,” “scripts,” “objects,” etc.), effectively detecting technical jargon and code context.

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

 take

-1.36

if

-1.32

 after

-1.26

 taking

-1.22

 giving

-1.18

but

-1.18

-1.17

ll

-1.16

cz

-1.16

-1.14

POSITIVE LOGITS

djang

1.49

 których

1.48

݀

1.45

 Consejos

1.44

 pomocą

1.40

Mvh

1.39

 Pristup

1.35

 אנשים

1.34

ň

1.33

colgante

1.32

Activations Density 0.148%