INDEX

Explanations

existing projects and assets

The neuron fires on mentions of concrete entities or structures—generic nouns naming existing objects or resources (e.g. “table,” “key,” “processes,” “site,” “measures”) in technical or procedural contexts.

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

 again

-1.32

sprogram

-1.29

なこと

-1.19

 arbe

-1.18

Ikus

-1.16

AndPassword

-1.16

 Kruse

-1.16

}_{

-1.15

 PLATFORM

-1.14

薷

-1.13

POSITIVE LOGITS

existing

1.75

 existing

1.69

 Existing

1.63

Existing

1.44

 bestehenden

1.38

现有

1.37

伱

1.25

既存

1.24

 existentes

1.23

 existente

1.23

Activations Density 0.034%