INDEX

Explanations

keyword class

The neuron fires on isolated numeric literals (numbers) in code snippets.

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

the

-1.89

最重要的

-1.75

how

-1.63

コミュ

-1.59

 später

-1.53

 does

-1.48

を中心

-1.48

 whose

-1.48

 necesitan

-1.45

 analogous

-1.45

POSITIVE LOGITS

1.72

)':

1.71

 monotony

1.62

丏

1.60

.+\

1.59

cccccccc

1.57

}>;

1.54

 Twelve

1.54

สำหรับ

1.53

 Czę

1.53

Activations Density 0.006%