INDEX

Explanations

lists or enumeration

The neuron fires on numeric tokens (e.g., counts, measurements, citation numbers) in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 потому

-1.47

 porque

-1.28

 поскольку

-1.25

因为

-1.24

 perché

-1.22

 BECAUSE

-1.19

 क्योंकि

-1.18

 octavo

-1.17

 mauva

-1.16

 protože

-1.16

POSITIVE LOGITS

 including

1.61

 from

1.59

or

1.58

 coupled

1.46

 which

1.43

 along

1.34

in

1.26

 combined

1.22

one

1.21

at

1.17

Activations Density 0.107%