INDEX

Explanations

make something [adjective]

The neuron fires strongly on content‐bearing nouns—especially longer, topic‐oriented or abstract nouns like “question,” “experience,” “plants,” or “process.”

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

 they

-1.59

 that

-1.57

but

-1.46

že

-1.44

merkt

-1.41

彼らは

-1.41

}$.

-1.35

They

-1.34

ilets

-1.34

leq

-1.28

POSITIVE LOGITS

 easier

1.56

’)

1.42

1.41

icône

1.31

狆

1.30

وروبا

1.30

 analogous

1.29

 mêmes

1.28

𐌸

1.27

 cintur

1.27

Activations Density 0.077%