INDEX

Explanations

age, height, years old

The neuron fires on proper names of people.

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

pos

0.62

les

0.61

ts

0.59

gr

0.58

 mini

0.57

 tapp

0.57

 cult

0.57

 offspring

0.57

 activity

0.57

 unle

0.56

POSITIVE LOGITS

８

1.00

۸

0.98

 eighty

0.97

７

0.97

９

0.93

۹

0.91

 nineteen

0.90

۷

0.90

 seventy

0.88

 seventeen

0.88

Activations Density 0.155%