INDEX

Explanations

The neuron fires on code‐documentation lines—especially docstrings or comments that describe classes, methods, or how to use them.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 wyniku

0.59

Untitled

0.57

mp

0.57

 темно

0.55

 pieces

0.55

 चीजें

0.55

readFileSync

0.55

̀

0.55

preventDefault

0.54

ต่างๆ

0.53

POSITIVE LOGITS

 class

1.63

class

1.53

 classe

1.34

 класс

1.28

 класса

1.25

クラス

1.24

 kelas

1.21

 clase

1.20

 classes

1.12

 klasse

1.11

Activations Density 0.020%