INDEX

Explanations

file path resource URLs

The neuron activates on numeric path segments and ID numbers—i.e. digit sequences embedded in URLs or file references.

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

んですよね

-1.41

 hacían

-1.35

toDouble

-1.33

 škola

-1.30

って何

-1.30

 insumos

-1.27

 soldadura

-1.27

 filós

-1.26

angnya

-1.25

ところに

-1.24

POSITIVE LOGITS

1.31

er

1.26

1.25

 haline

1.23

1.22

1.21

1.20

1.19

Activations Density 0.005%