INDEX

Explanations

personal possession and context

The neuron marks single capitalized words at the beginnings of sentences or text blocks.

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

Składniki

-1.51

塒

-1.44

 μαζί

-1.37

ﻒ

-1.34

 proposons

-1.33

 они

-1.27

 получите

-1.27

他们

-1.23

 Però

-1.23

 キャンプ

-1.21

POSITIVE LOGITS

1.71

to

1.63

 которы

1.53

}=\

1.44

 hardworking

1.41

1.39

return

1.38

__;

1.37

我也不

1.34

Activations Density 0.139%