INDEX

Explanations

Car models and brands

The neuron activates on proper‐noun product or manufacturer names (brand/model names).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 وكل

0.65

 כך

0.64

 globally

0.63

 hailing

0.61

 ومن

0.60

 upfront

0.59

नॉ

0.59

 日本

0.59

 글로벌

0.58

 नॉर्थ

0.57

POSITIVE LOGITS

Paged

0.55

transistors

0.51

 ignoreString

0.50

χο

0.50

рий

0.49

ലിയ

0.49

Temporal

0.46

takePhotoButton

0.46

 سالب

0.46

 medications

0.46

Activations Density 0.001%