INDEX

Explanations

named or defined entities

This neuron fires on the central subject noun of a passage (e.g. “library,” “constructor,” “class,” “herb,” “counter,” “column,” “protein,” “constraint,” “name,” “app,” etc.).

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

 MainAxisSize

-0.95

塍

-0.95

形で

-0.94

melhor

-0.94

きち

-0.93

 ziua

-0.90

GARET

-0.90

 trasladado

-0.89

 vernünf

-0.88

 llevado

-0.87

POSITIVE LOGITS

 named

1.45

“

1.42

of

1.42

 called

1.41

 which

1.36

1.32

である

1.31

1.29

\"

1.21

 namens

1.20

Activations Density 0.593%