Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 2644 in Layer 11

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 2.1285. Min Range: -2.1285

Max Act: 2.1285. Min Act: -0.0000

Data Index: 608665 (C4 (Web Text))

Max Activating Token Index: 699

Text #1

Max Range: 2.1285. Min Range: -2.1285

Max Act: 1.8348. Min Act: -0.0000

Data Index: 677505 (C4 (Web Text))

Max Activating Token Index: 580

Text #2

Max Range: 2.1285. Min Range: -2.1285

Max Act: 1.3392. Min Act: -0.0000

Data Index: 74396 (C4 (Web Text))

Max Activating Token Index: 341

Text #3

Max Range: 2.1285. Min Range: -2.1285

Max Act: 1.2167. Min Act: -0.0000

Data Index: 296147 (C4 (Web Text))

Max Activating Token Index: 617

Text #4

Max Range: 2.1285. Min Range: -2.1285

Max Act: 1.0522. Min Act: -0.0000

Data Index: 1469742 (Python Code)

Max Activating Token Index: 89

Text #5

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.9590. Min Act: -0.0000

Data Index: 113554 (C4 (Web Text))

Max Activating Token Index: 649

Text #6

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.9827. Min Act: -0.0000

Data Index: 51323 (C4 (Web Text))

Max Activating Token Index: 245

Text #7

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.9528. Min Act: -0.0000

Data Index: 186150 (C4 (Web Text))

Max Activating Token Index: 234

Text #8

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.8536. Min Act: -0.0000

Data Index: 1312128 (C4 (Web Text))

Max Activating Token Index: 828

Text #9

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.8955. Min Act: -0.0000

Data Index: 1573867 (Python Code)

Max Activating Token Index: 436

Text #10

Max Range: 2.1285. Min Range: -2.1285

Max Act: 1.0034. Min Act: -0.0000

Data Index: 634102 (C4 (Web Text))

Max Activating Token Index: 886

Text #11

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.9058. Min Act: -0.0000

Data Index: 1416678 (Python Code)

Max Activating Token Index: 315

Text #12

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.9010. Min Act: -0.0000

Data Index: 1295966 (C4 (Web Text))

Max Activating Token Index: 998

Text #13

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.8412. Min Act: -0.0000

Data Index: 875021 (C4 (Web Text))

Max Activating Token Index: 146

Text #14

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.9074. Min Act: -0.0000

Data Index: 415142 (C4 (Web Text))

Max Activating Token Index: 399

Text #15

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.7752. Min Act: -0.0000

Data Index: 735201 (C4 (Web Text))

Max Activating Token Index: 232

Text #16

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.7005. Min Act: -0.0000

Data Index: 1647327 (Python Code)

Max Activating Token Index: 459

Text #17

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.7868. Min Act: -0.0000

Data Index: 1644726 (Python Code)

Max Activating Token Index: 162

Text #18

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.7713. Min Act: -0.0000

Data Index: 234928 (C4 (Web Text))

Max Activating Token Index: 820

Text #19

Max Range: 2.1285. Min Range: -2.1285

Max Act: 0.8511. Min Act: -0.0000

Data Index: 240962 (C4 (Web Text))

Max Activating Token Index: 455