INDEX

Explanations

determiner preceding specific topic

The neuron is keyed to spotting Spanish definite articles (“el,” “la,” “las,” etc.) as they occur in formal headings or agenda‐item titles.

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

 Both

-1.59

"]:

-1.57

If

-1.55

);

-1.55

*}{

-1.48

-1.47

})-\

-1.47

 обязательно

-1.46

ítica

-1.45

 Like

-1.42

POSITIVE LOGITS

how

2.44

why

1.94

菝

1.91

豨

1.80

鲵

1.75

 pérd

1.70

Mucha

1.68

 milyen

1.62

盉

1.59

錞

1.54

Activations Density 0.118%