INDEX

Explanations

data structure sections

The neuron is triggered by numeric literals (digits and numbers) in code.

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

part

-2.06

褲

-1.71

most

-1.70

Ꮸ

-1.63

plus

-1.63

摭

-1.60

which

-1.55

 zahlreiche

-1.54

 '*':

-1.54

some

-1.52

POSITIVE LOGITS

我們

1.77

 aislados

1.70

已经

1.67

蹰

1.63

我们

1.59

 цвят

1.58

ність

1.57

同じ

1.55

лася

1.55

ος

1.54

Activations Density 0.000%