INDEX

Explanations

spaced, fast, or

This neuron detects words that express quantities, sizes, or measurements (e.g., “大,” “多,” “复杂,” “spaced”).

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

of

-2.03

 nėra

-2.03

 todėl

-1.96

ፁ

-1.88

焄

-1.81

 kurie

-1.79

 dabar

-1.78

硖

-1.74

着一

-1.72

彣

-1.70

POSITIVE LOGITS

 atau

2.03

 или

1.88

 большинстве

1.80

or

1.75

または

1.72

1.71

ish

1.70

或

1.67

 maioria

1.63

 یا

1.62

Activations Density 0.003%