INDEX

Explanations

accompanying figures or permission statements

This neuron detects tokens characteristic of code blocks—particularly indentation whitespace, punctuation like periods, 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

 unmarried

-0.71

"";

-0.68

 bread

-0.67

linea

-0.66

rej

-0.66

 Cauchy

-0.66

ัญ

-0.65

Ratio

-0.65

 squad

-0.65

进展

-0.65

POSITIVE LOGITS

resumen

0.75

体重

0.72

更有

0.71

τυ

0.70

Journalist

0.68

誰が

0.68

consejos

0.67

個人的に

0.66

Få

0.65

ची

0.65

Activations Density 0.064%