INDEX

Explanations

closing code structure

The neuron activates on numeric literal tokens (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

 máis

-1.01

كيف

-0.98

Մ

-0.97

gede

-0.97

على

-0.95

 SDLK

-0.95

來到

-0.94

الخ

-0.93

cuando

-0.89

 інтер

-0.88

POSITIVE LOGITS

of

1.13

 teaches

0.92

 principalmente

0.92

 предотвра

0.91

 yaitu

0.91

 interesantes

0.91

 many

0.90

 needed

0.90

 microfibra

0.89

was

0.88

Activations Density 0.028%