INDEX

Explanations

code block starts

The neuron fires on tokens involved in assignment statements—particularly user‐defined variable names (like map, i, se, root), the ‘=’ operator, and numeric literals.

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

𝐿

-1.38

Examples

-1.27

råd

-1.23

 véritables

-1.22

Please

-1.20

ungkap

-1.20

 défis

-1.19

Which

-1.17

 Übungen

-1.16

 provinsi

-1.15

POSITIVE LOGITS

 llamaba

1.39

but

1.37

 його

1.18

 will

1.14

its

1.05

 poche

1.05

 senhor

1.04

at

1.03

摂取

1.03

ייה

1.02

Activations Density 0.000%