INDEX

Explanations

two and both

The neuron detects mentions of duality—words or numeric expressions that refer to “both,” “two,” or paired items.

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

ραπε

-1.11

…？

-1.11

IANGLES

-1.09

tafogo

-1.04

 всю

-1.01

смартфон

-1.01

дован

-0.97

 deres

-0.95

それとも

-0.94

 from

-0.94

POSITIVE LOGITS

two

1.77

 halves

1.65

 ends

1.51

 sides

1.48

 beiden

1.38

 két

1.26

 dwie

1.23

 both

1.20

 δύο

1.13

 orejas

1.11

Activations Density 0.081%