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

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

Neuron 3972 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.5919. Min Act: -0.0000

Data Index: 1066282 (C4 (Web Text))

Max Activating Token Index: 687

Text #1

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.5012. Min Act: -0.0000

Data Index: 1180710 (C4 (Web Text))

Max Activating Token Index: 972

Text #2

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.4556. Min Act: -0.0000

Data Index: 604483 (C4 (Web Text))

Max Activating Token Index: 335

Text #3

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.3311. Min Act: -0.0000

Data Index: 1136319 (C4 (Web Text))

Max Activating Token Index: 775

Text #4

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.3466. Min Act: -0.0000

Data Index: 941543 (C4 (Web Text))

Max Activating Token Index: 665

Text #5

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.1129. Min Act: -0.0000

Data Index: 1397149 (Python Code)

Max Activating Token Index: 699

Text #6

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.1469. Min Act: -0.0000

Data Index: 525886 (C4 (Web Text))

Max Activating Token Index: 995

Text #7

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.1616. Min Act: -0.0000

Data Index: 1014623 (C4 (Web Text))

Max Activating Token Index: 496

Text #8

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.0748. Min Act: -0.0000

Data Index: 433626 (C4 (Web Text))

Max Activating Token Index: 713

Text #9

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.1160. Min Act: -0.0000

Data Index: 1395642 (Python Code)

Max Activating Token Index: 24

Text #10

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9396. Min Act: -0.0000

Data Index: 601683 (C4 (Web Text))

Max Activating Token Index: 992

Text #11

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.1176. Min Act: -0.0000

Data Index: 1478379 (Python Code)

Max Activating Token Index: 25

Text #12

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9842. Min Act: -0.0000

Data Index: 1128033 (C4 (Web Text))

Max Activating Token Index: 88

Text #13

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9755. Min Act: -0.0000

Data Index: 1258908 (C4 (Web Text))

Max Activating Token Index: 232

Text #14

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9719. Min Act: -0.0000

Data Index: 1065991 (C4 (Web Text))

Max Activating Token Index: 208

Text #15

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9151. Min Act: -0.0000

Data Index: 839343 (C4 (Web Text))

Max Activating Token Index: 358

Text #16

Max Range: 1.5919. Min Range: -1.5919

Max Act: 1.0849. Min Act: -0.0000

Data Index: 22713 (C4 (Web Text))

Max Activating Token Index: 210

Text #17

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.8822. Min Act: -0.0000

Data Index: 1379621 (Python Code)

Max Activating Token Index: 26

Text #18

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9435. Min Act: -0.0000

Data Index: 1434119 (Python Code)

Max Activating Token Index: 14

Text #19

Max Range: 1.5919. Min Range: -1.5919

Max Act: 0.9199. Min Act: -0.0000

Data Index: 1440433 (Python Code)

Max Activating Token Index: 612