INDEX

Explanations

keys in structured data

The neuron fires on identifiers naming data fields or model/table/variable properties in code (e.g., column or attribute 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

for

-1.33

 when

-1.13

in

-1.12

at

-1.11

 during

-1.08

Thanks

-1.08

 Segurança

-1.05

 Desenvolvimento

-1.05

Firstly

-1.03

 исключительно

-1.02

POSITIVE LOGITS

 årets

1.48

 kvinnor

1.38

潜力

1.38

ᯅ

1.19

erobic

1.18

 carreaux

1.16

 prochaines

1.15

ᜀ

1.15

瞟

1.15

 mortos

1.14

Activations Density 0.017%