INDEX

Explanations

digital twins and necessary requirements

The neuron fires strongly on technical or domain-specific content words (i.e. the key specialized terms in a passage).

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

diğim

-1.57

konstruktion

-1.51

Ekster

-1.51

پیوند

-1.49

因為

-1.48

 during

-1.48

เพราะ

-1.44

關於

-1.41

nhà

-1.41

 mohl

-1.40

POSITIVE LOGITS

1.59

 لاعب

1.42

 掃除

1.41

now

1.40

 リボン

1.38

 tibet

1.31

 jetta

1.30

 GENERATED

1.26

 palab

1.23

еты

1.22

Activations Density 0.161%