INDEX

Explanations

you/your

The neuron is looking for instructional or advisory language addressed to the reader (e.g. “you,” “your,” “help,” “tips,” “important,” etc., in how-to or “plan and prepare” 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

部份

-1.40

־ה

-1.38

 hitherto

-1.37

 tewas

-1.32

觅

-1.26

=>'

-1.22

Esperamos

-1.20

 dispara

-1.18

 invariably

-1.16

 einschließlich

-1.16

POSITIVE LOGITS

 your

2.27

you

1.57

leyebilirsiniz

1.45

你的

1.41

 také

1.34

 maaaring

1.34

 mężczy

1.28

おわりに

1.28

İN

1.27

bedaan

1.26

Activations Density 0.035%