INDEX

Explanations

only

The neuron fires on numeric quantity mentions—i.e. digits or counts (like “3 lifts,” “5 dice,” lab readings, etc.).

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

inosa

-1.11

or

-1.07

Insets

-1.05

ство

-1.03

 kolejny

-1.02

ledem

-1.02

ğından

-1.00

 from

-1.00

ằng

-1.00

 because

-0.98

POSITIVE LOGITS

 only

2.58

 فقط

2.27

 تنها

2.02

only

1.82

 solamente

1.80

 μόνο

1.80

 только

1.72

 alone

1.70

 един

1.69

 apenas

1.67

Activations Density 0.142%