Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 4127 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.9222. Min Act: -0.0000

Data Index: 190806 (C4 (Web Text))

Max Activating Token Index: 904

Text #1

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.7760. Min Act: -0.0000

Data Index: 1651888 (Python Code)

Max Activating Token Index: 968

Text #2

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.7784. Min Act: -0.0000

Data Index: 1063343 (C4 (Web Text))

Max Activating Token Index: 960

Text #3

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.7243. Min Act: -0.0000

Data Index: 1402317 (Python Code)

Max Activating Token Index: 789

Text #4

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.6580. Min Act: -0.0000

Data Index: 968821 (C4 (Web Text))

Max Activating Token Index: 433

Text #5

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.6684. Min Act: -0.0000

Data Index: 404757 (C4 (Web Text))

Max Activating Token Index: 222

Text #6

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5721. Min Act: -0.0000

Data Index: 1051771 (C4 (Web Text))

Max Activating Token Index: 661

Text #7

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5817. Min Act: -0.0000

Data Index: 880029 (C4 (Web Text))

Max Activating Token Index: 313

Text #8

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5796. Min Act: -0.0000

Data Index: 1554312 (Python Code)

Max Activating Token Index: 47

Text #9

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5930. Min Act: -0.0000

Data Index: 15794 (C4 (Web Text))

Max Activating Token Index: 904

Text #10

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5438. Min Act: -0.0000

Data Index: 743559 (C4 (Web Text))

Max Activating Token Index: 235

Text #11

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5602. Min Act: -0.0000

Data Index: 1654151 (Python Code)

Max Activating Token Index: 499

Text #12

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.6002. Min Act: -0.0000

Data Index: 1259756 (C4 (Web Text))

Max Activating Token Index: 914

Text #13

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.6047. Min Act: -0.0000

Data Index: 357088 (C4 (Web Text))

Max Activating Token Index: 360

Text #14

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5331. Min Act: -0.0000

Data Index: 941249 (C4 (Web Text))

Max Activating Token Index: 776

Text #15

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5454. Min Act: -0.0000

Data Index: 1553868 (Python Code)

Max Activating Token Index: 676

Text #16

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.4908. Min Act: -0.0000

Data Index: 1635191 (Python Code)

Max Activating Token Index: 980

Text #17

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5129. Min Act: -0.0000

Data Index: 547790 (C4 (Web Text))

Max Activating Token Index: 409

Text #18

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.4922. Min Act: -0.0000

Data Index: 1069621 (C4 (Web Text))

Max Activating Token Index: 606

Text #19

Max Range: 0.9222. Min Range: -0.9222

Max Act: 0.5392. Min Act: -0.0000

Data Index: 306766 (C4 (Web Text))

Max Activating Token Index: 785