Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 0 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 1.4363. Min Range: -1.4363

Max Act: 1.3509. Min Act: -0.0001

Data Index: 256798 (C4 (Web Text))

Max Activating Token Index: 336

Text #1

Max Range: 1.4363. Min Range: -1.4363

Max Act: 1.4363. Min Act: -0.0001

Data Index: 905783 (C4 (Web Text))

Max Activating Token Index: 157

Text #2

Max Range: 1.4363. Min Range: -1.4363

Max Act: 1.2184. Min Act: -0.0000

Data Index: 622169 (C4 (Web Text))

Max Activating Token Index: 309

Text #3

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.9618. Min Act: -0.0001

Data Index: 267195 (C4 (Web Text))

Max Activating Token Index: 928

Text #4

Max Range: 1.4363. Min Range: -1.4363

Max Act: 1.0298. Min Act: -0.0000

Data Index: 583451 (C4 (Web Text))

Max Activating Token Index: 733

Text #5

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.9990. Min Act: -0.0001

Data Index: 276554 (C4 (Web Text))

Max Activating Token Index: 727

Text #6

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.9691. Min Act: -0.0001

Data Index: 606404 (C4 (Web Text))

Max Activating Token Index: 590

Text #7

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.9007. Min Act: -0.0000

Data Index: 817011 (C4 (Web Text))

Max Activating Token Index: 81

Text #8

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.9686. Min Act: -0.0001

Data Index: 708370 (C4 (Web Text))

Max Activating Token Index: 765

Text #9

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8104. Min Act: -0.0001

Data Index: 1631818 (Python Code)

Max Activating Token Index: 194

Text #10

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8576. Min Act: -0.0000

Data Index: 16067 (C4 (Web Text))

Max Activating Token Index: 887

Text #11

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8337. Min Act: -0.0001

Data Index: 1307574 (C4 (Web Text))

Max Activating Token Index: 582

Text #12

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8558. Min Act: -0.0000

Data Index: 403262 (C4 (Web Text))

Max Activating Token Index: 609

Text #13

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8798. Min Act: -0.0000

Data Index: 161553 (C4 (Web Text))

Max Activating Token Index: 501

Text #14

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8807. Min Act: -0.0000

Data Index: 801064 (C4 (Web Text))

Max Activating Token Index: 796

Text #15

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8719. Min Act: -0.0000

Data Index: 907055 (C4 (Web Text))

Max Activating Token Index: 279

Text #16

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8154. Min Act: -0.0000

Data Index: 498849 (C4 (Web Text))

Max Activating Token Index: 438

Text #17

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.7603. Min Act: -0.0000

Data Index: 1155840 (C4 (Web Text))

Max Activating Token Index: 461

Text #18

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.8077. Min Act: -0.0001

Data Index: 153390 (C4 (Web Text))

Max Activating Token Index: 359

Text #19

Max Range: 1.4363. Min Range: -1.4363

Max Act: 0.7962. Min Act: -0.0001

Data Index: 1414792 (Python Code)

Max Activating Token Index: 589