INDEX

Explanations

/** or // comment lines

The neuron strongly activates on standalone numbers or other numeric tokens (years, IDs, codes) in the text.

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

 altı

-0.78

 hoger

-0.73

ernos

-0.72

isTrue

-0.70

 głę

-0.70

óta

-0.69

 arbejds

-0.68

🆚

-0.68

 wciąż

-0.68

 Melville

-0.68

POSITIVE LOGITS

 Mannheim

0.68

是以

0.66

 Parliamentary

0.66

 Business

0.65

 Hohen

0.64

 assassins

0.63

Servlet

0.62

createdBy

0.62

 installation

0.61

ºC

0.61

Activations Density 0.040%