INDEX

Explanations

tutoring, factory, document, dated

This neuron detects sequences of consecutive periods (ellipsis-style “…”).

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

 обзор

0.64

ashop

0.57

 формування

0.57

 श्रॉफ

0.57

注文

0.56

 изоли

0.56

 অজ্ঞাত

0.55

 orphaned

0.55

 దృష్టి

0.54

 lineare

0.54

POSITIVE LOGITS

.....

0.60

 .......

0.57

.......

0.56

.........

0.56

…..

0.55

 .....

0.55

……..

0.55

 Falle

0.55

ਿਸ

0.55

 ……..

0.55

Activations Density 0.000%