INDEX

Explanations

quoted phrases within text

The neuron fires on text enclosed in quotation marks, i.e. direct quotations.

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

的法

-1.32

our

-1.27

government

-1.24

 dobře

-1.20

yuu

-1.20

的血

-1.18

governmental

-1.15

 különböző

-1.13

的身影

-1.13

 zcela

-1.12

POSITIVE LOGITS

 then

1.68

 тогда

1.34

Then

1.30

 THEN

1.30

แล้ว

1.29

 Then

1.28

 затем

1.24

特别是

1.23

 если

1.23

 burada

1.20

Activations Density 0.034%