Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 77 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1407. Min Act: -0.0001

Data Index: 65278 (C4 (Web Text))

Max Activating Token Index: 668

Text #1

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1386. Min Act: -0.0002

Data Index: 532887 (C4 (Web Text))

Max Activating Token Index: 30

Text #2

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1379. Min Act: -0.0001

Data Index: 702721 (C4 (Web Text))

Max Activating Token Index: 445

Text #3

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1300. Min Act: -0.0002

Data Index: 669743 (C4 (Web Text))

Max Activating Token Index: 550

Text #4

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1279. Min Act: -0.0002

Data Index: 588557 (C4 (Web Text))

Max Activating Token Index: 380

Text #5

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1194. Min Act: -0.0002

Data Index: 1220648 (C4 (Web Text))

Max Activating Token Index: 697

Text #6

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1220. Min Act: -0.0002

Data Index: 242609 (C4 (Web Text))

Max Activating Token Index: 346

Text #7

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1218. Min Act: -0.0002

Data Index: 256346 (C4 (Web Text))

Max Activating Token Index: 170

Text #8

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1123. Min Act: -0.0002

Data Index: 1192383 (C4 (Web Text))

Max Activating Token Index: 27

Text #9

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1084. Min Act: -0.0002

Data Index: 444000 (C4 (Web Text))

Max Activating Token Index: 126

Text #10

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1068. Min Act: -0.0002

Data Index: 740484 (C4 (Web Text))

Max Activating Token Index: 369

Text #11

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1095. Min Act: -0.0002

Data Index: 745226 (C4 (Web Text))

Max Activating Token Index: 467

Text #12

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1123. Min Act: -0.0002

Data Index: 527586 (C4 (Web Text))

Max Activating Token Index: 82

Text #13

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1040. Min Act: -0.0002

Data Index: 1134497 (C4 (Web Text))

Max Activating Token Index: 790

Text #14

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1088. Min Act: -0.0002

Data Index: 719930 (C4 (Web Text))

Max Activating Token Index: 691

Text #15

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1007. Min Act: -0.0002

Data Index: 982963 (C4 (Web Text))

Max Activating Token Index: 36

Text #16

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1049. Min Act: -0.0002

Data Index: 1348530 (C4 (Web Text))

Max Activating Token Index: 133

Text #17

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1036. Min Act: -0.0002

Data Index: 310091 (C4 (Web Text))

Max Activating Token Index: 103

Text #18

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1080. Min Act: -0.0002

Data Index: 1354405 (C4 (Web Text))

Max Activating Token Index: 252

Text #19

Max Range: 0.1407. Min Range: -0.1407

Max Act: 0.1025. Min Act: -0.0002

Data Index: 1262891 (C4 (Web Text))

Max Activating Token Index: 218