INDEX

Explanations

code includes and function calls

The neuron fires on programming‐style tokens—especially file paths, filenames, and code identifiers (e.g. method or variable names).

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

Odpo

-1.75

Specifik

-1.74

Podob

-1.60

ınır

-1.46

 internetu

-1.39

Zdra

-1.36

 triko

-1.34

Kde

-1.34

inize

-1.32

 тщательно

-1.31

POSITIVE LOGITS

с

2.28

С

2.13

с

2.03

С

2.02

Д

1.91

А

1.83

 по

1.81

К

1.77

и

1.76

р

1.76

Activations Density 0.127%